The Cost of Non- Europe in the Sharing Economy

The Cost of Non-

Europe in the

Sharing Economy

Economic, Social and Legal

Challenges and Opportunities

STUDY EPRS | European Parliamentary Research Service Author: Pierre Goudin European Added Value Unit PE 558.777- January 2016

The Cost of Non-Europe in the Sharing Economy

On 27 January 2015, the Coordinators of the European Parliament's Committee on Internal Market and Consumer Protection (IMCO) agreed to request a European Added Value assessment on the opportunities and challenges of the sharing economy.

This paper has been drawn up by the European Added Value Unit of the Directorate for Impact Assessment and European Added Value, within the European Parliament's Directorate- General for Parliamentary Research Services. Its aim is to help improve understanding of the subject matter by providing evidence of the specific benefits that could be achieved through European action.

This assessment builds on expert research commissioned specifically for the purpose and provided on the one hand by Europe Economics and on the other by the European Institute for Public Affairs (EIPA).

Abstract

This 'Cost of Non-Europe' study examines the current economic, social and legal state of play regarding the sharing economy in the European Union, and identifies the cost of the lack of further European action in this field.

The assessment of existing EU and national legislation confirms that there are still significant implementation gaps and areas of poor economic performance. The subsequent examination of areas where it was believed that an economic potential exists highlighted that substantial barriers remain, hindering the achievement of the goals set out in the existing legislation. Moreover, some issues are not or are insufficiently addressed (e.g. status of workers employed by sharing economy service providers). Consequently, more European action would be necessary to achieve the full economic potential of the sharing economy. In doing so, policymakers should seek to ensure an adequate balance between creative freedom for business and the necessary regulatory protection.

This research estimates the potential economic gain linked with a better use of capacities (otherwise under-used) as a result of the sharing economy is €572 billion in annual consumption across the EU-28. This figure should nevertheless be considered with caution; substantial barriers prevent the full benefits from being realised, and could reduce the value of potential increased use to up to €18 billion in the shorter-term and up to €134 billion in the medium and longer term, depending on the scale of regulatory obstacles.

EPRS | European Parliamentary Research Service

AUTHOR Pierre Goudin, European Added Value Unit

ABOUT THE PUBLISHER This paper has been drawn up by the European Added Value Unit of the Directorate for Impact Assessment and European Added Value, within the Directorate-General for Parliamentary Research Services of the Secretariat of the European Parliament.

To contact the Unit, please e-mail [email protected]

LANGUAGE VERSIONS Original: EN Translations: DE, FR

This document is available on the internet at: www.europarl.europa.eu/thinktank

DISCLAIMER The content of this document is the sole responsibility of the author and any opinions expressed therein do not necessarily represent the official position of the European Parliament. It is addressed to the Members and staff of the EP for their parliamentary work. Reproduction and translation for non-commercial purposes are authorised, provided the source is acknowledged and the European Parliament is given prior notice and sent a copy.

Manuscript completed in January 2016. Brussels © European Union, 2016.

PE 558.777 ISBN 978-92-823-8586-9 doi: 10.2861/26238 QA-01-16-059-EN-N

PE 558.777 2

Table of Contents

Executive Summary ...... 5

Introduction ...... 7

1. The sharing economy: a global state of play...... 9

1.1. A brief historical overview ...... 9

1.2. A definition of the sharing economy ...... 10

1.3. Drivers and issues ...... 11 1.3.1. Technology ...... 12 1.3.2. Evolving economic behaviours ...... 12 1.3.3. Social and societal factors ...... 13 1.3.4. An economic dimension ...... 13 1.3.5. A social dimension: ...... 15 1.3.6. An environmental dimension: ...... 17 1.4. The need for more or new regulation vs a laissez-faire approach ...... 17

2. The sharing economy in the EU: synthesis of the findings ...... 19

2.1. The economic and social potential of the sharing economy...... 19

2.2. Obstacles and barriers, and their economic value ...... 21

2.3. The existing EU regulatory framework: efficiency and limits ...... 23 2.3.1. Existing framework ...... 23 2.3.2. Remaining sharing economy issues ...... 26

3. Recommendations ...... 28

3.1. Priority issues directly relevant for the sharing economy ...... 28 3.1.1. Defining digital platforms ...... 28 3.1.2. Improving regulations applied to sharing economy platforms...... 28 3.1.3. Mitigating 'social exclusion' ...... 29 3.1.4. Dealing with potential market power of sharing economy platforms ..... 29 3.1.5. Applying labour market regulation to sharing economy platforms ...... 30

3.2. Other initiatives ...... 31

Annex I – The Cost of non-Europe in the sharing economy Research paper by Europe Economics ...... 33

Annex II – Cost of non-Europe in the sharing economy: legal aspects Briefing Paper by European Institute of Public Administration ...... 127

PE 558.777 3

Note on methodology

Costs of Non-Europe (CoNE) reports are designed to study the possibilities for economic benefits and the achievement of a 'public good' through common action at EU level. They attempt to identify policy areas which can benefit from deeper EU integration, where the added value of action at EU level is potentially significant.

This Cost of Non-Europe report principally deals with the opportunities and challenges of the sharing economy within the EU market and seeks to address the question of the added value of an EU-level response to the challenges identified. It notably analyses existing legislation, identifies specific gaps where legislation at European level could be beneficial and quantifies the costs borne by various stakeholders under the current regulatory framework. ,. The report is not intended to be exhaustive, but rather to provide concrete examples of the costs incurred by citizens and business in various areas, as well as to outline the wider economic and social impact of the identified gaps on the functioning of the internal market.

The report relies both on quantitative and qualitative analysis of data. Specific case studies are selected to illustrate the main findings.

PE 558.777 4

Executive Summary

In recent years, a major evolution has begun to reshape the capitalist economy. Today, no-one can say with certainty to what extent the 'sharing economy' will change the economic landscape. Yet, the speed, dynamism and scale of the change seem to point to a substantial long-term trend. What is at work here is a transition from traditional individual ownership of most assets towards accessibility-based economic models, which can be observed across a wide and increasing variety of markets.

Two trends can be observed in the evolution of this rental-like model. Firstly, technological progress allows this new business model to spread to more and more markets and become more and more convenient and flexible. Secondly, there is a shift to a peer-to-peer accessibility- based business model, centred on companies that operate through an online platform or marketplace that connects consumers owning certain assets and skills with consumers in temporary need of them.

There seems to be no consensus at EU level on either the name or the definition of these new economic models: while some institutions have chosen to call the phenomenon the 'collaborative economy', others prefer to refer to the 'sharing economy'. This is the case, in particular, of the European Parliament, and for the purpose of this study, the sharing economy will be defined as 'The use of digital platforms or portals to reduce the scale for viable hiring transactions or viable participation in consumer hiring markets (i.e. 'sharing' in the sense of hiring an asset) and thereby reduce the extent to which assets are under-utilised.'

The scope of this study thus includes activities providing access to the following goods and services: accommodation, transport, consumer durables, labour and human capital, intellectual property and professional services. It should be noted that finance falls outside this scope of this study.

In any case, the sharing economy emerges as a complex issue. Beyond the problem of the definition itself, this type of activities raises a number of important and controversial questions in the economic, as well as in the social and environmental fields. Additionally, it remains difficult to predict its development and future with certainty. The dazzling start of this new economy may give hope for a brighter tomorrow; some indeed do not hesitate to predict tremendous growth rates and a windfall of benefits. Others are wary of the speed and magnitude of the success of the sharing economy, expressing doubts at the long-term nature of this development and warning of a disappointing slowdown.

In order to provide a comprehensive picture, this study addresses these issues at three levels: - firstly, it attempts to evaluate the economic and social potential of the sharing economy in the European Union. - secondly, it seeks to identify whether there are obstacles or barriers preventing the sharing economy from reaching its full potential, and if so, the economic impact thereof. - Finally, it analyses the effectiveness of the existing EU regulatory framework in promoting and overseeing such emerging business models.

Its main findings are as follows:

1) The nature of the sharing economy is likely to change over time as it grows in scope and matures. Consumers are likely to benefit from lower prices and an increased quality of services. Providers will enjoy new economic opportunities, but may not benefit from the other advantages associated with traditional employment. Manufacturers may need to

PE 558.777 5

adapt to a market, in which fewer goods, but of higher value, are consumed. Moreover, other policy priorities are also likely to be affected by the sharing economy: its growth could eventually lead to a reduction in income and wealth inequality. At the same time, however, its development could potentially trigger the creation of new forms of 'social exclusion', such as the exclusion of an individual/provider from the sharing economy business due to e.g. poor ratings.

2) The potential economic gain linked with a better use of capacities (otherwise under-used) as a result of the sharing economy is estimated at €572 billion. This amount is theoretical, in so far as substantial barriers currently prevent the full benefits from being realised.

3) The current regulatory framework would in principle allow the sharing economy to continue growing, but will not enable the best feasible results to be met in the medium term. Maximising the potential of the sharing economy would thus require new initiatives at European Union level.

The set of recommendations listed at the end of this study suggests the additional steps which ought to be taken at the EU level in order to achieve the full economic potential of the sharing economy. In doing so, policy-makers should seek to ensure an adequate balance between creative freedom for business and the necessary regulatory protections.

PE 558.777 6

Introduction

In recent years, a major evolution – probably even a revolution – has begun to fundamentally change the capitalist economy in most countries around the world. Today, no-one can say with certainty to what extent the 'sharing economy' will change the economic landscape. Yet, the speed, the dynamism and the scale of the mutation seem to point to a substantial long term trend. What is at work here is a transition from traditional individual ownership of most assets towards accessibility-based economic models. Such a transition can be observed across a wide and increasing variety of markets: delivery services, home troubleshooting, transport, cooking, housework, locksmiths, plumbers, hotel booking, travel, banking, car rental, to name but a few. In the conventional situation, consumers would buy products and become the owners; in an accessibility-based system, consumers pay for temporary access rights to a product. Clearly this type of business has been conducted for several decades already (probably even longer for some goods), for instance in the form of car rental services in business-to-consumer (B2C) markets and outsourcing in business-to-business (B2B) markets.

This conventional business model, however, is now subject to radical change. Two trends can be observed in the evolution of this rental-like model. Firstly, technological progress allows the business model to spread to more markets and become more convenient and flexible. An example of this is the Spotify music streaming service that provides consumers access to an estimated over thirteen million music tracks, conveniently through their smartphone, tablet or computer. Another example is the Car2Go car rental company: which provides members with flexible and local access to individual mobility through a large quantity of rental cars that are distributed across European cities. These are typical examples of accessibility based business models in the B2C market.

A second trend constitutes a shift to peer-to-peer accessibility-based business models. In the conventional model, companies provide access for consumers to company owned property; in peer-to-peer models, companies facilitate access for consumers to consumer owned property or skills and competencies. Most of these companies function through an online platform or marketplace that connects consumers. Thus, they link people that own certain assets and skills with consumers in temporary need of them. These companies can facilitate peer-to-peer markets for potentially all products or services owned by consumers. This business model might become particularly disruptive to conventional rental solutions for mobility, accommodation, catering and other services: it is indeed able to serve the same needs at a significantly lower price. Moreover, it empowers consumers to capitalise on their property and skills, providing them with an opportunity for micro-entrepreneurship and lowering the total cost of ownership. The partially disruptive specificity of this business is also the fact that, for some services, labour and workforce themselves become a good, which can be traded on the market.

There are several macro-economic factors driving the growth of the sharing economy. One such factor is certainly decreasing consumer trust in the corporate world – especially among young people – as a result of the financial and economic crisis. In addition, unemployment rates have risen and the purchasing power of consumers has fallen. Therefore, citizens are seeking ways to earn or save money; which is why consumers are currently more receptive to peer-to-peer business models centred on consumer needs both as a potential supplier and buyer. Furthermore, the technology required for hosting an online peer-to-peer market has, in recent

PE 558.777 7

years, become available at a more reasonable cost. Finally, environmental concerns also explain the increased use of the sharing economy. Sharing underutilised assets, reducing waste and promoting the transition to a more environmentally friendly economy are initiatives which have gained increasing support especially amongst younger people.

Beyond the strictly macroeconomic factors, the success of the sharing economy is also linked to generation, to culture and to development of a sharing mentality. The younger generation is more familiar with new technologies and masters them better. This generation has also suffered most from the economic crisis and is therefore more suspicious of established systems and open to alternative solutions. Thus, it is sensitive to the stated intention of the sharing economy which aims to facilitate daily life by involving the end user in the production process of the service.

As a result, the potential of the sharing economy is significant, with annual growth exceeding 25%; in some sectors it could even reach 63% by 2025.1

This growing popularity of the sharing economy clearly has implications: - Traditional, incumbent companies risk the possible loss of a significant share of their market; - Regulatory bodies face new challenges, since innovation is outstripping their ability to regulate the industry effectively.

Rules suitable to facilitate and coordinate business-to-Business or business-to-consumer transactions are not always applicable to the newly created Consumer-to-Consumer market. Likewise, rules governing the physical world may not always work effectively when applied to the intersection between the digital and the physical sphere: sometimes, they no longer preserve a level playing field for traditional economic actors and newcomers. Within the EU, the approach chosen to respond to this phenomenon varies significantly from one Member State to the other. As a result, the various legal frameworks are quite different and difficult to compare, thus contributing to the fragmentation of the Single Market.

Against this background, the European Parliament has decided to commission a Cost of Non- Europe report on the opportunities and challenges of the sharing economy. This report provides an overview of its economic potential and the main challenges to be addressed. The report suggests a series of common initiatives at European level in order to enable the sharing economy to achieve its full potential and to promote a flexible environment for innovation. In this respect, policy-makers should seek to ensure an adequate balance between creative freedom for business and the necessary regulatory protections.

1 PWC (2015), Press release.

PE 558.777 8

1. The sharing economy: a global state of play

1.1. A brief historical overview

The sharing economy refers to a business model that actually belongs to a 'family' with multiple organisational schemes: some of them are very simple – barter – other much more sophisticated – online exchange platforms, based on complex algorithmic software.

The appearance of sharing economy schemes in historical and geographical terms varies from one model to another: bartering goes back to ancient times and is practiced all around the world, while trading platforms have only emerged in the last few years – in connection with the development of the internet and smartphones – and if their expansion is global, it assumes the presence of communities of critical size and an enabling environment (accessibility) to be economically viable. Between these two opposite examples, many other forms of sharing economy – based on pooling resources – have been tested over time and still work: cooperatives, mutual societies, associations and foundations, tontines.

These different models have common elements that are more or less similar to each other; however, they come from different 'philosophies' and have neither the same economic rationale nor the same purpose. Some of them are not profit-based businesses – they fall into the sphere of the social economy; others are for-profit companies but their organisation and governance comply with ethical goals – they could be classified as social entrepreneurship. Others choose a form of entirely for-profit business: this is the case, mainly, for exchange platforms, created mostly in the form of start-ups and whose sharing element lies not in their organisation, but in the object of their activity.

Today, the sharing economy is a notion that sometimes tends to compete – or, at least, to be 2 placed in parallel – with the notion of collaborative economy, popularised in recent years to describe this new and growing economic model. In any case, there seems to be no consensus at EU level on the definition of the sharing economy. The European Commission prefers to use the expression 'collaborative economy', defined as 'a complex ecosystem of on-demand services and temporary use of assets based on exchanges via online platforms'.3 The other EU institutions do use the expression 'sharing economy'. The European Parliament refers to it in its resolutions of 9 September 20154 and 29 October 2015,5 and defines it as: 'a new socio-economic model that has taken off thanks to the technological revolution, with the internet connecting people through online platforms on which transactions involving goods and services can be conducted securely and transparently'. The European Economic and Social Committee also referred to the sharing economy in its Opinion of 21 January 2014.6 Finally, the Committee of the Regions has

2 R. Botsman and R. Rogers (2010), 'What's Mine is Yours : The rise of collaborative consumption', Harper Business. 3 COM (2015) 550 final. 4 European Parliament resolution of 9 September 2015 on the implementation of the 2011 White Paper on Transport: taking stock and the way forward towards sustainable mobility (2015/2005(INI)). 5 European Parliament Resolution of 29 October 2015 on new challenges and concepts for the promotion of tourism in Europe (2014/2241(INI)). 6 European Economic and Social Committee (2014), 'Collaborative or participatory consumption, a sustainability model for the 21st century'.

PE 558.777 9

recently published an opinion,7 where it argues in favour of the need to distinguish between the different forms of sharing economy; it calls for a coordinated approach between the European Commission and the Member States in order to enable successful sharing economy initiatives to spread easily across EU borders.

The multiple nuances that exist between the various concepts just mentioned, which are sometimes more than shades, underline therefore the importance of and the need for a precise definition of what is meant by sharing economy.

1.2. A definition of the sharing economy

As the purpose of this study is not, however, to state a new theoretical or semantic work on the concept of the sharing economy, the approach retained is rather to set out a pragmatic and efficient definition. Its goal will be to avoid artificial distinctions between economically similar activities and prove to be amenable to economic analysis and the identification of potential policy measures.

Some established definitions already exist, among which:

− a very broad definition is suggested by The People Who Share, a campaigning group promoting the sharing economy: 'The sharing economy is a socio-economic ecosystem built around the sharing of human and physical resources. It includes the shared creation, production, distribution, trade and consumption of goods and services by different people and organisations. ... The sharing economy encompasses the following aspects: swapping, exchanging, collective purchasing, collaborative consumption, shared ownership, shared value, co-operatives, co-creation, recycling, upcycling, redistribution, trading used goods, renting, borrowing, lending, subscription based models, peer-to-peer, collaborative economy, circular economy, pay-as-you-use economy, wikinomics, peer-to-peer lending, micro financing, micro-entrepreneurship, social media, the Mesh, social enterprise, futurology, crowdfunding, crowdsourcing, cradle-to-cradle, open source, open data, user generated content.'; − in contrast, a quite narrow definition from a more academic world8 attempts to define the sector more closely by arguing a) it should only include consumer-to-consumer transactions, not business-to-consumer transactions, b) the sharing economy should be understood to only include transactions where consumers provide temporary access to a good, not the permanent transfer of ownership of the good, c) it should only include transactions regarding physical assets.

Such definitions are doubtless a helpful starting point. Ultimately, they are either too broad or too narrow, which does not enable an understanding of the sharing economy as a properly circumscribed economic phenomenon.

7 'The local and regional dimension of the sharing economy', Committee of Regions Opinion Number: CDR 2698/2015., 3-4 December 2015. 8 K. Frenken, T. Meelen, M. Arets & P. Van de Glind (2015), 'Smarter regulation for the sharing economy', The Guardian, blog.

PE 558.777 10

Therefore, the whole analysis of this study is based on a new and tailored definition of the sharing economy. In its research for the European Parliament, Europe Economics defines the sharing economy as:

The use of digital platforms or portals to reduce the scale for viable hiring transactions or viable participation in consumer hiring markets (i.e. 'sharing' in the sense of hiring an asset) and thereby reduce the extent to which assets are under-utilised.

Such a definition has several features that can also be seen as real assets: − it defines the sharing economy by a combination of two elements: first, the sharing economy is considered as an opportunity to reduce the extent to which assets are under-utilised, by employing a rental model; second, it is made possible by technological breakthroughs which have reduced transaction costs and increased the extent to which sharing is now accessible to many more people. Taking advantage of that opportunity to extend rental markets constitutes the sharing economy; − to some extent, it reflects what some people have described as broader categories including the sharing economy, such as the Collaborative Economy, or includes sectors which others have defined as similar to, but not a part of, the sharing economy, e.g. the Product-Service Economy; − it focuses on consumer markets (i.e. peer-to-peer or business-to-consumer markets), as opposed to pure business-to-business markets (which appear to be a separate phenomenon that might have quite different economic impacts and policy implications).

Finally, the definition of the sharing economy used could include activity on platforms providing access to the following goods and services: − accommodation; − transport; − consumer durables; − labour and human capital; and − intellectual property.

It might be noted that the definition could also include finance, however finance falls outside the scope of this study. The role of the sharing economy in finance might be quite different and the regulatory considerations are unique; it is therefore set aside here and should be considered in further research.

1.3. Drivers and issues

The sharing economy emerges as a complex issue not only with regard to problems arising from its definition, which results, depending on the selected wording, in substantial differences in the scope of this activity, the nature of players participating, the policies applying, and the solutions for which it calls.

The sharing economy also seems complex because: − Firstly, it raises a number of important and controversial questions in the economic, as well as in the social or environmental fields;

PE 558.777 11

− Secondly, it remains difficult to predict its development and future with certainty. The dazzling start of this new economy may give hope for a brighter tomorrow; some indeed do not hesitate to predict tremendous growth rates and a windfall of benefits. Others are wary of the speed and magnitude of the success of the sharing economy, expressing doubts as to the long term nature of this development and warning of a disappointing slowdown.

At this stage, only the factors favouring the emergence of the sharing economy, boosting its growth and generating the involvement of a constantly increasing number of people are well known and identified. The rise of the sharing economy is driven and enabled by converging changes in some markets and areas:

1.3.1. Technology

Technology is a key driver and booster for the sharing economy: such a change would not have been possible without the development of the internet, mobile devices and digital platforms that facilitate individual access to many services and play an intermediary role in linking the supplier and the user of these services.

The use of these technological advances has probably also been reinforced by the parallel development of social networks – themselves helped by technological innovation. These networks, by developing the notion of communities, have encouraged the development of relationships and interactions specific to these communities; they have established them as full actors in the economic field. New needs and new demands specific to these communities have appeared, as they have gradually revealed and imposed themselves in their capacity to act as stakeholders, in prescribing trends, and as lobbyists (particularly through the use of evaluation systems). Finally, advances in technology have also played a role in the growth of the sharing economy by allowing paperless financial transactions: online or mobile payment systems to develop hand in hand with the rise of e-commerce and digital platforms; they allow ordinary individuals to achieve modest peer-to-peer economic transactions, sometimes single transactions, which would not have been possible previously due to a lack of adequate support and back office facilities.

1.3.2. Evolving economic behaviours

Evolving economic behaviours also play in favour of the sharing economy: the effects of the financial and economic crisis since 2008 have significantly and durably impacted household purchasing power; many people seek both to make savings on their expenditure and to find supplementary income. Today, studies increasingly point out that many consumer goods are only used a fraction of the time or for only a part of their abilities. The conjunction of the crisis with the realisation of this stockpile of underutilised assets and development of information technology (IT) applications enabling creation of 'bespoke' services has finally promoted the rise of a new form of economy. This economy is based on a streamlined utilisation of a hitherto neglected or unknown economic wealth, individualised supply of services (and individualisation of price fixing), and on shifting boundaries between economic actors. In this context, the user of an asset may the same day – or the next – become the supplier of another good, although some users are involved in the definition and production of the service they purchased. Finally, it can not be denied that some environmental awareness leads to challenge:

PE 558.777 12

the wasteful behaviour of consumer society and the resulting fight against waste can be faced by more and better use of any asset. All of these economic adjustments are converging.

1.3.3. Social and societal factors

Finally, social and societal factors are at work to foster the emergence, development and lasting quality of a modern sharing economy. First and foremost, the growth of the world population and the increasing concentration of people in cities facilitates production of local services, which are at the heart of the sharing economy. Similarly, the densification of the population in a limited number of places creates a favourable context for developing communities, whose role and importance have been mentioned above. Subsequently, the concentration of people, specifically in urban locations, creates new needs, specific to these forms of organisation: this is the case, in particular, in terms of mobility. It is no coincidence that some of the most iconic and successful platforms for the sharing economy are platforms supplying and promoting motor vehicle rental or sharing, or other means of transport. Finally, environmental concerns influence the increased use of the sharing economy. It was stated earlier that such concerns result in part from an economic rationality – the desire to fully take advantage of an asset, once acquired; they can also be explained by motivations directly related to sustainable development: mutualising assets, sharing their use, processing for reuse, are all ways to prepare a transition to a greener economy.

Through factors on which the sharing economy relied to secure its expansion, one can observe issues which are at stake, but also questions about its future in the medium and long term. This economy is based on three dimensions that are central to the major questions of modern society.

1.3.4. An economic dimension

Clearly, the sharing economy has not really invented a new economic model: the movement is part of a story, and relies on a series of practices that it has improved, modernised and optimised; however, its growth, helped by new technologies and the continuing impact of the crisis, also reflects a collective will to act and spend differently. It gradually imposes an economic model that arises as a complementary model – if not a true competitor – to the traditional capitalist model.

No one can truly say today how this new model will evolve, nor what its future will be. Will it deliver on its promises; will it remain just a complementary economy; or is it doomed to be 'a flash in the pan'? In his latest book,9 Jeremy Rifkin predicts the greatest success for the sharing economy. Joseph Stiglitz, 2015 Nobel Laureate in Economics, in turn appears much more cautious, even sceptical. In an interview with Le Monde,10 he questions 'the innovations of recent years'. 'For now, Facebook, Airbnb, the collaborative economy does not generate productivity gains as powerful as those of the industrial revolution, and we do not know how to measure what they inject in gross domestic product.'

So, today, it remains difficult to quantify the economic contribution of the sharing economy, even if it claims to be presented as a proper sector per se. According to a study by the European Commission, the revenue generated by the sharing economy for individuals who use it to

9 J. Rifkin (2014), 'The Zero marginal cost society'; Palgrave Macmillan publishers. 10 J. Stiglitz (2015); Le Monde, 'L'Union européenne est en train de détruire son avenir'.

PE 558.777 13

supplement their income reached US$3.5 billion worldwide in 2013. Already, the turnover of this economic sector is estimated at €20 billion and, according again to the European Commission, the market for the sharing economy could eventually reach US$110 billion.11 It is undeniable that this economy is experiencing almost unprecedented growth and is becoming customary at an incredible speed: in France alone, according to a recent study,12 while 19% of consumers reported use of this type in 2013, twice as many considered it in 2014, and 60% in 2015. Furthermore, the phenomenon is not just about small innovative start-ups: large companies have already realised that this new economic model affects many sectors of activity, and that this new consumption model challenges the whole trade chain. Big traditional groups are thus now investing in the sharing economy and buying start-ups that enrich their production processes and their supply range.

However, the sharing economy continues to raise doubts among some economists, who emphasise a peculiar paradox: the digital economy, although valued in billions of euros, affords no real solution in terms of stable and sustainable growth. The sharing economy does not properly create assets by itself, and most of its actors do not create added value; peer-to-peer services between consumers, in particular, generate little VAT. Therefore, economic production is not improved. Most of the new self-employed workforce who belong to the sharing economy are often less productive – in a macroeconomic sense – than if they were part of a traditional company and sectors in which this economic model thrives are generally low value-added and low-skilled. As a result, productivity gains from technological progress seem to be offset or negated by the extension of a scattered and unskilled workforce.

Can we therefore talk about a bubble effect? High-tech companies created since the beginning of the 21st Century do indeed advertise sometimes extravagant valuations, not directly related to their tangible assets: the stock market valuation of Airbnb now exceeds that of the Accor Group, the world leader in hotels; when Airbnb employs 600 people worldwide, compared with 180 000 employees for the Accor group, which operates 3 700 hotels. The assets, on which such potential overvaluations in these new areas are made, may indeed be questioned. What will happen if the bubble bursts?

Furthermore, the development of the sharing economy is not without any consequence for the traditional economy, to which it not only adds, but more often substitutes. Do the activities and jobs created actually balance or even exceed the number of companies forced to close, and the resulting redundancies? Again, it is very difficult today to measure competition and substitution effects, which remain controversial: firstly, as has already been stated, because the sharing economy is still a recent phenomenon that is difficult to quantify with certainty; and because some of the mechanisms at work are incredibly complex. The example of the automotive sector is, in this respect, a very illustrative case study: in the study attached, Europe Economics estimates that if the European car fleet was 100% utilised through the sharing economy, it would be possible to decrease the fleet of 200 million vehicles. What would be the undeniably dramatic impact on the European automotive industry and its thousands of employees? Another study,13 however, points out that additional income generated by renting and sharing an under-utilised vehicle via a digital platform mostly leads its owner to then

11http://ec.europa.eu/enterprise/policies/innovation/policy/business-innovation-observatory/files/case- studies/12-she-accessibility-based-business-models-for-peer-to-peer-markets_en.pdf. 12 Enquête Fevad/CSA sur les perspectives d'achat sur Internet en 2015. 13 L. Belot (2015), 'La déconnexion des élites: comment Internet dérange l'ordre établi'; Editions Les Arènes.

PE 558.777 14

invest in a car of a superior category. Similarly, a more intensive use of a vehicle causes a shortening of its life cycle, and therefore the need to replace it sooner. Are these positive economic behaviours (from an automotive industry perspective) sufficient to offset the negative effects resulting from a drastic reduction in the car fleet? Some economists fear the ultimate consequence of the sharing economy would be that there is nothing left to share.

In any case, conventional businesses, challenged by those belonging to the sharing economy, will have to adapt, even if only because of how the value of a product is determined is about to be questioned and reset: in the field of the sharing economy, this value will be increasingly assessed and evaluated over the long term by a community of consumers who have shared or exchanged the product to which this value is attached. For brand owners' businesses, this means that they will now have to address a new audience – consumer communities – and review their customer relations policy. Facing such a significant phenomenon, companies will have to reorganise internally, according to Altimeter Group.14 More and more services will be crowdsourced. The status and role of the client will themselves become more undefined, since they will become a stakeholder in the company's project, by participating in consumer communities that help companies to improve their products and services.

Beyond performance, viability and the sustainability of its model, the sharing economy still poses other problems of an economic nature. The first major question raised by this emerging economic sphere is related to taxation. Today, in many cases, start-ups in the sharing economy are 'free-riders' in their behaviour towards the social systems in which they thrive. Existing tax regimes were not designed for activities or goods such as those of the sharing economy, and many of the businesses in that sector feel they do not have to submit to taxation. However, this situation, a fortiori when facing an economy that is growing strongly, results in a growing shortfall for public finances; it generates cascading effects, in particular for the welfare state model. Furthermore, the absence of taxation can lead to some unfair competition in relation to professionals subject to payroll taxes, and often forced to respect strict rules specific to their business (security, traceability.).

A similar problem arises for insurance, the traditional system in force not being adapted to the nature of services supplied by the sharing economy, nor that of the economic relations established between its customers. For instance, traditional insurers have not yet developed horizontal insurance products suitable for service providers participating in multiple platforms. In some cases only, timid and limited attempts have emerged, constituting merely early examples of self-regulation in this field.

1.3.5. A social dimension:

The sharing economy fosters specific economic models; it redefines the concept of work and thereby has an impact on the labour market. Thus, it encourages all actors in the social field to reflect on the meaning and place that the salary system has in our society, and thereby, to rethink social protection, historically linked to the salary system.

Indeed, most digital platforms, because of their specific – and relatively new – organisation, lead to a redefinition of the very concept of work: instead of depending on the authority of an

14 J. Owyang & al. (2013), 'The collaborative economy'; Altimeter Group.

PE 558.777 15

entrepreneur, individual work is now intermediated by computers. Thereby, the work becomes subject to an exchange on the market and to continuous trading, as for a raw material. The employment contract is replaced by the sale of a service whose amount and prices vary in real time, according to supply and demand.

It is quite clear that the current organisation of labour markets does not match the business model of the sharing economy. Protest movements by workers depending on these companies have emerged: some refuse the self-employed status and claim to be employees. Indeed, many workers in the sharing economy are not employees and therefore do not benefit from a guaranteed or minimum remuneration, since they are theoretically free to supplement their income with another application, another platform. However, workers denounce working conditions that prevent them from multiplying contractors (exclusivity requirement imposed although unwritten, long hours and onerous performance indicators), which render them 'disguised' employees. Today, it is the courts that have to resolve these disputes, based on increasingly inadequate laws. The time will soon come, however, when the law will have to adapt. The question is how, and how far?

The likely evolution of the rules on work will notably have to address the risks of exclusion generated by the functioning of the sharing economy, for a whole range of workers. The Europe Economics study attached addresses this issue, even though it defines social exclusion in a very limited way. It should be noted however, more broadly, that according to another study,15 'on demand' workers are overwhelmingly male (72.7%), young (70% are aged between 18 and 34 years) and single (65%). In the USA, this new labour market organisation is known as the '1099 economy', referring to the form number corresponding to independent contracts. An economy that so massively promotes a single profile category creates some risk for employment levels – and beyond, on the ability itself to be employed – for the rest of the workforce in any specific market; the social consequences that may result warrant vigilance as to the evolution of employment, and a readiness to legislate if necessary.

Social protection is another field that the specific organisation of the sharing economy disrupts, or even subverts. A refusal or 'neglect' by many players in this economy to pay social contributions may eventually jeopardize the proper functioning and effectiveness of our social protection systems. However, such behaviours are indicative of the growing inadequacy of these systems in relation to the new business reality of the sharing economy. Such an economy is not based on a collective and centralised structure, but on a multitude of 'contractors', independent from each other. The issue may ultimately arise, of a move to a model where contributory obligations and social protection are no longer linked to employment status (salary system) but the individual.

In any case, the fragility of many of the companies in the sharing economy – many are born, many die, perhaps even faster than in the traditional economy – and the weakening status of their workers will lead to a reflection on better career security. Again, the right balance between the need for regulation and a concern not to stifle the innovation and expansion that characterise today's sharing economy. It cannot be excluded, however, that when companies operating in this sphere have reached critical mass and a sustainable development pattern, they develop their own solutions, including insurance, which respond to the expectations and social needs of their workers.

15 Requests for startups (2015); 'The 2015 1099 economy workforce report'

PE 558.777 16

1.3.6. An environmental dimension:

The growth of the sharing economy raises hope among proponents of a sustainable development that respects our environment and efficient use of resources. The model on which it is based – pooling and sharing of goods and services, increased use and optimal reuse of these assets, with priority given to local trade patterns –seems indeed well in line with the characteristics and criteria of a greener economy; and would even be indispensable for an ecological transition. As noted by the European Economic and Social Committee,16 'Collaborative or participatory consumption could prove resilient in the current economic and financial climate, and provide a response to the growing uncertainties caused by the economic crisis. It could also represent an opportunity to get back on track towards sustainable economic, social and human development in an environmentally-friendly way'.

At present, according to ADEME (French Agency for Environment and Energy Control), 17 40% of freezers and refrigerators are replaced while still in working order; the situation is the same for 25% of dishwashers and 14% of washing machines. As to the technical lifespan of a mobile phone, it is estimated at more than ten years, but users tend to change device every two and a half years on average. With the sharing economy, the use-life of these asset increases, and their use is maximised. In principle, this situation leads to less production and reduced resource harvesting and waste generation, for the same level of service.

Nevertheless, 'The environmental balance is less obvious than it seems,' notes IDDRI (Institute for Sustainable Development and International Relations) in their study.18 Some considerations may explain this paradox: − firstly, different consumption does not necessarily mean consuming less: the money saved in the context of the sharing economy – for example by choosing to use carpooling – may be later spent on air travel, for example. This is known as the 'rebound effect'; − furthermore, it appears that practices related to the sharing economy may ultimately prove to be ambivalent: should the use of carpooling enjoys massive success, it could encourage governments to limit their investments in public transport.

Therefore, the sharing economy may be presented as a tool for ecological transition only if it meets a number of conditions, such as the durability of the goods or a change in habits in relation to consumption. For all these issues, consumers may be mobilised, but public authorities also have a role to play, and are limited by the necessity to follow the rules.

1.4. The need for more or new regulation vs a laissez-faire approach

16 European Economic and Social Committee (2014), 'Collaborative or participatory consumption, a sustainability model for the 21st century' 17 ADEME (2012); 'Etude sur la durée de vie des équipements électriques et électroniques' 18 D. Demailly & A-S. Novel (2014), 'Économie du partage : enjeux et opportunités pour la transition écologique'; IDDRI

PE 558.777 17

In response to the growing importance of the sharing economy, several Member States have already looked for legal solutions. Generally speaking, there is a shared concern by all, which is the necessity to adapt to technological innovations while ensuring respect for fair competition. This balance has not yet been found, as the varying responses demonstrate. The situation in the Member States ranges from outright prohibition to a more friendly approach and wavers between more regulation and, on the contrary, simplification measures.

These existing legal provisions are quite sparse and call rationally for a common legal framework at EU level, covering some – if not most – activities included in the sharing economy. Two types of approach may be applied: government control (or top-down government regulation) or bottom-up regulation (or self-regulation through reputation). Best practices deployed at Member States level indicate that a mix of both approaches will presumably be needed.

PE 558.777 18

2. The sharing economy in the EU: synthesis of the findings

The definition of the sharing economy settled, the attached studies had to answer three major questions:

− what is the economic and social potential of the sharing economy in the European Union? − are there obstacles or barriers which prevent the sharing economy from reaching its full potential, and if so, what is their economic value? − how effective is the existing EU regulatory framework in promoting and overseeing such emerging new business models?

2.1. The economic and social potential of the sharing economy

Regarding the first issue (what is the economic and social potential of the sharing economy in the European Union?), the key findings were the following:

a) The nature of the sharing economy is likely to change over time as its scope and scale grow, and as it matures

In fact, the potential of the sharing economy will depend on how it develops, with the growth or contraction of existing platforms and the creation of new platforms or changes in existing platforms. The sharing economy is likely to extend to new markets (particularly including those relating to marketing and other transaction costs, education and health and markets, in which sharing economy platforms are not currently able to attain critical mass, but where they might expand with growing scale and/or technical changes, e.g. from urban to suburban or even rural areas), while peer-to-peer transactions are likely to decline as proportion of the sharing economy. In many markets, the peer-to-peer element seems likely to be a feature of a transition to the sharing economy: the trend is that, after growing strongly, the P2P share of the market as a whole peaks at around 25% (when ownership is around 50%) and then begins to decline. Ultimately, the market is dominated by B2C rental. More broadly, the sharing economy's future development might be dominated by new potential business models in existing sectors expanding out of niches in a continuous process over time, or by an expansion into new sectors entirely (changing the implications for policy).

b) Consumers will tend to benefit from lower prices and increased quality of services, along with a reduction in the 'lumpiness' of their consumption and the ability to satisfy more diverse preferences over time.

Prices can be lower for consumers using sharing economy platforms, for three reasons: increased utilisation, increased supply and/or lower costs. Quality might be improved through three channels: enhanced transparency (through public ratings systems); increased competition leading to improvements in new and existing providers; new innovations reflected more quickly in the capital stock. As to the diversity of choice, it is likely to grow in importance with

PE 558.777 19

the development of the market, as smaller niches will represent more viable opportunities for sharing economy providers, and may also become more important as firms offer services through those platforms rather than peers. It represents an alternative to mass customisation, satisfying greater diversity in tastes over time as opposed to greater diversity in tastes between consumers.

c) Providers may enjoy new economic opportunities, but may not enjoy access to the other benefits associated with traditional employment.

Many providers will enjoy higher aggregate earnings in the sharing economy than they would have otherwise, as they receive additional earnings alongside their prior income. However, the impact of growth in the sharing economy on average earnings in a given sector is likely to depend on the source of the reduction in consumer costs: this impact is likely to be an increase in aggregate labour earnings and an increase in individual earnings for new entrants to a sector, without which they would not enter. It also should be noted that some benefits might not be available to those working as self-employed providers in the sharing economy, including paid holiday, paid sick leave, employer pension contributions, maternity and paternity leave and employment protection. However it is important to bear in mind that this situation may be a distinctive feature of self-employment, rather than of the sharing economy itself.

d) Competing providers outside the sharing economy may face increased competition in service markets, but the sharing economy will generally not increase the pressure on scarce resources such as land for development or road space.

There are broadly three types of market participant which at some level compete with the sharing economy and might therefore be affected by its development: − those providing substitute goods and services; − those who have other uses for scarce assets; − those manufacturing goods for ownership.

Existing services could clearly be affected by the reduction of barriers to entry and therefore more competition in the markets in which they operate.

e) Manufacturers may need to adapt to a market in which fewer, but higher value, goods are consumed. Other policy priorities will also be affected, with a reduction in income and wealth inequality but the potential to create new forms of social exclusion. The use of electronic payments and digital platforms makes a significant improvement in tax compliance possible.

Manufacturers have the potential to be significantly affected by the growth of the sharing economy. The most direct effect might be that if assets are used more efficiently, there might be less demand and therefore volumes might fall significantly. That might be a challenge for established manufacturers, leading to excess capacity. Manufacturers might gain, however, if they are able to either deliver a higher-value product or offer associated services.

PE 558.777 20

There are concerns that the sharing economy might exacerbate inequality. However, to the extent that the welfare gains from the sharing economy accrue most to those with limited access to valuable assets at present, it is likely to reduce inequality. Beyond that, the sharing economy might reduce the salience of wealth inequality, by reducing the degree to which either wealth, or the ability to borrow, is necessary to access valuable assets.

In some markets, the sharing economy might reduce social exclusion by increasing access to goods and services. If the ability to access these markets becomes increasingly essential, however, and platforms and market participants are extremely risk averse, then it might be difficult for those who do not appear reliable to those participants to establish themselves in the market.

f) The obstacle-free theoretical maximum potential reduction in under-utilisation associated with the sharing economy amounts to €572 billion, although that is subject to a number of (in some cases quite fundamental) barriers (Cf. next section).

− The value of the under-utilisation of labour across the EU-28 is €309 billion on this estimate. − The average under-utilisation of accommodation for the EU-28 is 3%, i.e. around €35 billion per year. − Consumption of cars is around €500 per person, or €254 billion in total, across the EU- 28. To take the lower end of the estimated range at 60%, we can therefore estimate a potential under-utilisation of €152 billion in annual consumption. − Other sectors imply an aggregate underutilisation across the EU-28 estimated between €38 and €76 billion, according to different scenarios.

2.2. Obstacles and barriers, and their economic value

Concerning the second issue (are there obstacles or barriers which prevent the sharing economy from reaching its full potential, and if so, what is their economic value?), the key findings were as follows: a) The need for a certain level of digital access and skill is currently a significant obstacle to the sharing economy but one expected to decline in importance rapidly.

If smartphone penetration continues to rise to 90% or more, which is expected to happen in some Member States by 2018, then digital access and skills will become a less significant obstacle to the growth of the sharing economy. b) Physical barriers to participation in the sharing economy are significant but may be overcome by new business models.

There are a number of geographical and other physical barriers which might limit the development of the sharing economy, including:

PE 558.777 21

− low population density: most sharing economy platforms are currently focused on serving customers in urban areas. This may reflect in part that there are advantages to sharing economy services that are less salient in suburban or rural areas; − high costs of transport for sharing: there are some goods and services where transport costs make sharing economy services prohibitive. Many consumer durables, for example, are very heavy and could not readily be shared.

c) Consumer preferences for ownership are a significant obstacle to sharing, but can be reduced in importance as features of sharing economy markets today, such as product scarcity risk and a lack of diversity in products on offer, become less pronounced over time.

A pure preference for ownership might, at least in part, be a temporary result of people being accustomed to owning certain assets. If it becomes more practical and therefore more common to hire those assets then, over time, such a preference for ownership might decline. Furthermore, to the extent the sharing economy extends into new markets where product scarcity risk is particularly serious (e.g. human health services), platforms might need to develop new means of assuring consumers that assets will be available when needed. Platforms might create some kind of reserve of providers who are paid to act as a provider of last resort in the event that others are unavailable, creating a capacity market. d) In some economies, labour market obstacles, e.g. skills mismatches, will inhibit the growth of the sharing economy.

There might indeed be a range of reasons why those who are unemployed or under-employed might be unable to take up work in the sharing economy, including low mobility, sticky wage demands, technical or social skills mismatches.

e) The need to establish trust is a key challenge for the growth of the sharing economy, but one that platforms can meet over time in a range of different ways.

Consumers renting goods or buying services in the sharing economy need to trust that the service will be delivered to a reasonable standard at the expected price, or that they will get proper compensation if it does not, and that their safety and security will be maintained. There are several strategies by which platforms might try to create that trust, including insurance, prior scrutiny before participants in the market start using the platform and ratings once those participants have started using the platform.

f) Tax and other policy choices not intended to affect the sharing economy might still affect its growth in each economy.

Tax policy might, in some cases, inhibit the development of the sharing economy in two ways:

− high taxes on the returns from establishing sharing economy platforms might mean that fewer platforms are established. This might have a number of effects: reducing competition between platforms; hindering the development of potential European

PE 558.777 22

competitors to US platforms; and potentially slowing the development of new business models. − high or complex taxes (creating a compliance burden) might discourage providers and lead to a reduction in supply. g) Regulation can deter sharing economy growth through outright bans, regulatory costs which deter self-employment, regulatory costs which deter marginal transactions or through inconsistencies and idiosyncrasies in intellectual property rules.

There are a number of Member States in which sharing economy applications have been banned outright because they are not compliant with regulatory structures applied to established providers, or are subject to regulatory requirements which most sharing economy providers are not able or expected to meet. These bans can have effects that cross national borders. Firstly, they inhibit the development of services that cross Member State borders which might thereby encourage other business between Member States. Secondly, they potentially favour local providers of booking services. Thirdly, some of the rulings prohibit consumers using the services in other Member States, where they are clearly legal.

2.3. The existing EU regulatory framework: efficiency and limits

It should be noted that the sharing economy is not immune to the rules and policies implemented by both the Member States and the European Union. Some of its players suggest that this is not the case, because many existing rules appear unsuited to the new economic model of the sharing economy. However, even though no rule may seem to exist, relations between individuals are regulated, at least in civil law.

A regulatory framework already exists, however, particularly at European level, which applies wholly, or in part, to the sharing economy; both measures addressing digital services in particular and those addressing broader regulatory policy. Examination of this framework will demonstrate the framework's effectiveness and limitations, the latter relying on existing provisions but also on topics not covered. Although the sharing economy is and would be able to grow under the current framework, it does not mean that the current framework will produce the best feasible results from the sharing economy in the medium term. The following statement calls therefore for new initiatives from the European Union; as stated in a report by the European Economic and Social Committee, 'Because collaborative consumption represents a substantial economic, social and cultural shift, the Commission should remove any obstacles to these activities at the European level, establishing a regulatory framework that offers the sector certainty for the long term.' Likewise, research for the European Commission has argued that 'for the sharing economy in general, it would be beneficial to have specific legislation for sharing initiatives in various industries', in order to avoid a 'lack of clarity because existing legislation does not cover certain activities and transactions' or the possibility that 'legislation developed for conventional industries is wrongfully applied to markets in the sharing economy'.

2.3.1. Existing framework

Given its wide-ranging nature the sharing economy is affected by an equally wide range of EU policy. There are two broad relevant areas: policy which affects sharing economy platforms

PE 558.777 23

such as digital services; and policy which affects sharing economy services, through their effects on the general regulation of the services which are provided through those platforms.

Policy affecting sharing economy platforms as digital services includes:

− The E-Commerce Directive (2000/31/EC), which defines that information society services are subject to the law of the Member State in which the service provider is established and that Member States cannot restrict incoming services. This principle might be extended in the context of the sharing economy by making it easier for consumers to use platforms with which they are familiar in other Member States. − The Commission has recently reviewed the legal framework on the protection of personal data, aiming to modernise the legal system, strengthen individual rights and improve the clarity and coherence of the rules. The European Parliament and the Council reached an agreement on the Data Protection Reform. The relevant instrument regarding the sharing economy is the 'General Data Protection Regulation' which aims to a) enable citizens to exercise effectively their right to personal data protection (TFEU, Article 16(1)), and b) modernise and unify rules so that business makes the most of the Digital Single Market. While the Regulation seems to offer answers to some of the concerns raised by the sharing economy, its final text is not yet available.19 However, the principle should remain that 'personal data can only be gathered legally under strict conditions, for a legitimate purpose.' This may affect certain proposals for changes to ratings systems over time. − The Digital Single Market Strategy proposed by the Commission includes proposals to construct a new regulatory framework for online platforms, in part through a new Internal Market Strategy and e-commerce framework. As a part of the strategy, the Commission also proposes to address 'a number of concerns over the growing market power of some platforms'.

Relevant policy affecting the markets in which sharing economy providers compete includes:

− The Services Directive (2006/123/EC), which aims to ensure that customers benefit from stronger rights, higher quality services and enhanced information about providers, while businesses benefit from easier establishment, easier provision of cross- border services, and simplified procedures and formalities. Under the Internal Market Strategy for Goods and Services (CWP 2015), the aim is to 'deliver further integration and improve mutual recognition in key industrial and services sectors'. Providers offering their services through sharing economy platforms could clearly fit within this principle. − The Directive on Consumer Rights (2011/83/EC) regulates contracts between consumers and traders. This generally aims to strike a balance between robust consumer protections and ensuring businesses can remain competitive. Its application to sharing economy platforms should provide for price transparency with rules against hidden charges, and requiring total costs to be made clear. Rules against pre-ticked boxes could, however, affect opted-in benefits for sharing economy providers. Moreover, with regards to consumer protection in the sharing economy, it is unsure whether the Directive would be applicable to all types of sharing economy platforms.

19 See: http://ec.europa.eu/justice/data-protection/reform/index_en.htm (last consulted on 12/01/2015).

PE 558.777 24

While the Directive applies to both sales and service contracts, some areas are out of scope. − The Working Time Directive (2003/88/EC) provides for limits on working time. Member States may potentially derogate many limitations on working time for those with 'autonomous decision-taking powers'; which has often been applied to self- employed workers, including those offering their services through sharing economy platforms. − The Employment Information Directive (91/553/EEC), defining how workers should be told about terms and conditions; the Citizens Rights Directive (2004/38/EC), which gives workers the right to move freely and work anywhere in the EU; and other components of EU labour law create a framework in which sharing economy providers will work, although in some cases transactions will take the form of a contract between businesses (with one of the parties a self-employed contractor), rather than one between worker and employee. − The Professional Qualifications Directive (2005/36/EC) aims to facilitate the mobility of labour within the EU by allowing those qualified in one Member State to work in their profession in another Member State without repeating the qualification process. Automatic recognition in key professions is made possible by minimum training requirements, which evaluations have shown need to be updated over time to remain relevant and sufficient. This might provide a precedent for some form of common standard. − More recently, in the Communication on 'Upgrading the Single Market: more opportunities for people and business' of 28 October 2015,20 the Commission said that it would provide guidance on how EU law applies to collaborative economy business models in 2016, rather than strictly regulating the issue. In particular, it will draw upon national, European and international existing legislation to identify best practices, analyse how regulatory gaps need to be filled, and monitor development.

Topics which are ongoing or under review relevant to the future of the sharing economy

− In its Communication on 'a Digital Single Market Strategy for Europe',21 the Commission decided to assess 'the role of platforms, including in the sharing economy, and of online intermediaries'. As part of this assessment, it has launched an online public consultation from 24 September to 30 December 201522 monitored by the European Commission's Directorate-General for Communication networks, Content and Technology, and Directorate-General for Internal Market, Industry Entrepreneurship and SMEs. − In parallel, the Commission launched two studies, one by the Directorate-General for Mobility and Transport on 'passenger transport by taxi, hire car and ridesharing in the EU' and another by the Directorate-General for Justice and Consumers on 'consumer issues in the sharing economy', both expected in the second quarter of 2016.

20 COM(2015) 550 final. 21 See footnote n°21 22 https://ec.europa.eu/digital-agenda/en/news/public-consultation-regulatory-environment-platforms- online-intermediaries-data-and-cloud (last consulted on 01/11/2015).

PE 558.777 25

− Finally, The European Commission recently adopted a proposal23 for a directive 'on certain aspects concerning contracts for the supply of digital content', that could be relevant to some transactions carried out in the context of the sharing economy.

2.3.2. Remaining sharing economy issues

Despite the considerable existing volume of policy applicable to the sharing economy, significant issues remain that require consideration. A new policy addressing these issues might allow regulatory objectives (such as tax collection and consumer protection) to be achieved at lower cost.

Achieving regulatory objectives at higher cost than necessary Broadly speaking, it has been quite difficult for authorities to regulate many of the services where sharing economy platforms are currently growing, in particular because they often concern interactions between a large number of relatively small businesses. Authorities have to take on more of the burden of ensuring consumer safety and other regulatory compliance than would otherwise be the norm. This means significant costs both for the regulating authorities and for the regulated providers, which it may be possible to reduce with the growth of sharing economy platforms.

Market segmentation and restriction While regulation of online services is conducted with strong rules to ensure a Single Market through the E-Commerce Directive, the offline goods and services offered through sharing economy platforms are the subject of varied and inconsistent Member State or local regulation. This regulation impedes the development of a Single Market in sharing economy platforms, as these platforms are not able to operate in certain Member States. This limits competition among providers and could therefore lead to higher prices for consumers. It also means that the size of the market for platforms in Europe is limited.

Not making use of platform data Tax collection, in particular, is a perennial problem in many of the sectors in which sharing economy platforms operate. Member States are not making full use of the potential of the growth of these platforms and the data that they routinely collect on transactions. While any collection of data would need to be done sensitively, in order not to violate the principles underpinning data collection rules, it represents a significant opportunity.

Potential social exclusion The sharing economy has the potential to encompass a significant portion of economic and social life; and this might create a danger of a new (and potentially rather comprehensive) form of social exclusion. Users of certain sharing economy platforms whose reputational ratings fall below key thresholds are excluded from the platform. Those so excluded may find it impossible

23 COM(2015) 634 final

PE 558.777 26

to re-enter the platform to rebuild their reputation, because they cannot update their scores once they are excluded.

There is also some risk that users could become excluded maliciously or frivolously. These risks should be addressed from a public policy perspective.

Do sharing economy platforms naturally tend to become monopolies? A successful platform, particularly in markets with significant network effects, may tend towards becoming the sole (or overwhelming majority) player in providing the marketplace for some particular sharing economy activity. A natural concern, therefore, is that as sole (or overwhelming majority) players, sharing economy platforms might become monopolies. This could affect both consumers and providers.

Should all sharing economy service providers be employees of platforms? One of the key features of services provided via sharing economy platforms is that the service providers would, outside the sharing economy, naturally be employees and have additional security and benefits. Relevant regulations that might be linked to employment include: − minimum wage and working time regulations; − responsibility for safety and other working conditions; − employer-mandated welfare provision such as sick leave, healthcare or pensions; and − the administrative element of tax.

The most difficult of these issues is where public policy uses duties imposed upon employers as a mechanism for the provision of social insurance through welfare provision. In Europe it is much less common than in the USA for employer duties to be a key mechanism for healthcare provision, but that still leaves issues such as pensions or sick leave unaddressed.

Consumer protection At last, as stated in the abovementioned opinion of the European Economic and Social Committee, 'some measures should support, complement and monitor the consumer protection policy implemented by the Member States: a legal and tax framework for the activities covered by sharing consumption by setting down and regulating, where appropriate, aspects such as liability, insurance, rights of use, rights and obligations and, where appropriate, the removal of any restrictions and disguised barriers to intra-Community trade and any distortion of legislation.'

PE 558.777 27

3. Recommendations

The set of recommendations listed below addresses the following key question: what additional steps should be taken at EU level to realise the economic potential of the sharing economy, while continuing to balance creative freedom for business with necessary regulatory protections? Two options are selected:

− focusing in detail on a short number of priority issues; − enumerating a list of other topics that are more or less directly connected with the sharing economy but that should also be addressed by European policies.

3.1. Priority issues directly relevant for the sharing economy

3.1.1. Defining digital platforms

It appears important to establish clear criteria to determine to which legal category digital platforms belong: should they be considered information society services, or industry-specific businesses? It is worth noting that there is a pending case before the Court of Justice on the matter; the ruling is due in the course of 2016.

Even if digital platforms were to be considered as information society services, the current regulatory framework, in particular the e-Commerce Directive, would still not be fit for purpose. The legal regime needs to be updated in order to embrace changes relating to the sharing economy. The creation of a hybrid category for information society services, with a more balanced legal regime than that currently used, is therefore worth considering.

This modernisation of the legal framework is all the more important as the difference in the applicable legislation for offline and online services clearly has detrimental effects. The difference in the regime is perceived as encouraging unfair competition and companies are thus simply resorting to an online platform to avoid fulfilling their obligations.

Furthermore, the EU should provide guidelines on the threshold between what constitutes a professional activity exercised on a sharing economy platform and what does not. Looking at the best practices/examples analysed in different European cities, some elements could help the legislator to set a common level playing field. These include time and space limits, as well as income thresholds.

3.1.2. Improving regulations applied to sharing economy platforms

Shared economy platforms collect substantial amounts of data. On the one hand, this situation provides the opportunity to improve tax compliance at a lower cost. At the same time, making use of platform data could also help in decision-making and in accomplishing more general regulatory objectives. Such an approach would consist of delegating regulatory functions to the providers in a number of areas (registration and identification of market participants, confirming tax receipts, collecting taxes.). A common principle, however, would be to set reasonable regulatory requirements and then seek the most cost-efficient procedure by which

PE 558.777 28

the platform can ensure providers (and consumers) meet those requirements. This approach also implies the removal of quantitative restrictions and/or giving platforms an assurance that such restrictions will not be introduced.

In order to apply this at the EU level, there are broadly three alternative options: − defining a common objective, − establishing a common method, − setting common rules.

The last option of setting common rules ought to be favoured. Yet, this process should be carried out progressively, with clear focus on a fairly narrow set of sectors, for which there is an obvious need for a new legal framework (e.g. passenger transport). In doing so, the legislator should prevent over-regulating start-ups, which need room to manoeuvre to innovate and grow, and therefore concentrate on already well-established sectors (e.g. apartment rental). In order to prevent the sharing economy from being curtailed or driven along pre-determined lines, common rules should be set with a view to a possible rolling back of legislation in the medium term. This would also enable Member State authorities to accompany the development of the shared economy in an appropriate and flexible manner. Finally, the legislator should comply with the principles of subsidiarity and proportionality.

However, it would be worth also considering the role self-regulation could play. As highlighted by Europe Economics, a solution might lie in the outsourcing of certain legislative and control functions to the platforms.

3.1.3. Mitigating 'social exclusion'

The exclusion of an individual from the sharing economy due to poor ratings can have substantial consequences. Since errors are not excluded (i.e. due to malicious ratings or to market participants' inability to rehabilitate themselves after genuine lapses), the level of error, which would be perceived to be tolerable from a platform perspective (or too expensive to be worth eliminating), might not be considered as socially desirable from a public policy perspective.

Therefore, new measures seem justified in support of the rehabilitation of those excluded from platforms, including the prospective establishment of community platforms for that purpose. However, this should not occur through the regulation of still evolving financial ratings systems. Possible options in addressing this issue are the following alternatives:

− tolerating a degree of social exclusion (laissez faire approach), − establishing a right to a reputational Year Zero, − regulating reputational scoring so that only socially desirable exclusions occur, − creating community platforms where reputation can be rebuilt.

Providing a way out of exclusion from the shared economy to grant individuals a second chance would certainly secure higher social acceptance (even if some degree of social exclusion is unavoidable). The least intrusive form of such public intervention would be the establishment

PE 558.777 29

of community rehabilitation platforms, which would enable reputations in the sharing economy to be rebuilt.

In this respect, sharing economy reputation scoring systems would need to be more developed before an appropriate regulatory standard for fairness could be defined and enforced. Ultimately however, the creation of a legal framework on the principles and functioning of reputational rating systems would be crucial in boosting the trust consumers have in sharing economy platforms.

3.1.4. Dealing with the potential market power of sharing economy platforms

A general presumption that all sharing economy platforms will develop a dominant position is not founded. For the time being at least, the application of existing competition rules should ensure the required dynamism of digital markets.

Some sharing economy platforms may however become monopolies. Possible solutions to address this phenomenon could consist in:

− relying upon market forces and innovation to undermine market power, − developing the Single Market so as to maximise the size of the market, creating the greatest scope for multiple platforms, − using existing competition rules to identify instances of market power and specific appropriate interventions, − Treating sharing economy platforms in a manner analogous to regulated utilities.

A possible combination of competition and contestability/appeal possibilities – potentially fostered and facilitated by the extension of the Single Market – should be adequate to curtail market power. Therefore, to the extent that competition, contestability and the extension of the Single Market do not undermine market power, referring to competition authorities would remain the most appropriate step, before any economic regulation is called upon in this respect.

3.1.5. Applying labour market regulation to sharing economy platforms

Labour market regulations should not be altered to specifically include sharing economy providers. People working for providers should be allowed to remain self-employed, and platforms should be enabled (and in some cases, encouraged) to develop their own means of supplying other benefits besides cash remuneration. A remaining issue is whether providers should be considered employees of platforms. Exploratory avenues to answer this question could be to:

− mandate that all sharing economy service providers are platform employees, − create a new employment status of 'sharing economy service provider', − avoid extensive roles for employers in public welfare provision, − encourage or facilitate platforms in developing their own user benefits,

PE 558.777 30

− extend insurance and other financial markets.

As a conclusion, the most appropriate move would consist of including sharing economy service providers in the scope of the general rules applicable to self-employment. This option would be preferable to the other possibilities outlined, which suggest either assimilating workers for sharing economy service providers to employees or creating a new 'sharing economy service provider' employment status. The best complementary approach would be to allow (and in some cases potentially encourage) platforms to develop their own benefits options that would compete with the insurance products users could obtain for themselves.

3.2. Other initiatives

Next to the abovementioned areas, and directly related to the sharing economy, a number of fields exist where policy adaptations might be required to contribute to maximising the potential of the sharing economy:

− data protection rules: the principle according to which 'personal data can only be gathered legally under strict conditions, for a legitimate purpose' should be maintained; − manufacturing sectors: existing efforts to create a framework for growth in the manufacturing areas, e.g. the CARS 2020 Action Plan in the automotive sector, should take account of the possible impact of the sharing economy; − planning: in certain areas, such as transport infrastructure, planning should take the growth potential of the sharing economy into consideration in their calculation in terms of volume and need. − intellectual property rules: ongoing reform might need to take account of the increased importance of a cross-border hiring model, which may render geo-blocking and related restrictions superfluous.

PE 558.777 31

PE 558.777 32

Annex I

The Cost of non-Europe in the sharing economy

Research paper by Europe Economics

Abstract

The sharing economy can be understood as the use of digital platforms or portals to reduce the scale for viable hiring transactions or viable participation in consumer hiring markets and thereby reduce the extent to which assets are under-utilised. The notional obstacle-free potential to reduce under-utilisation is found to be €572 billion in annual consumption across the EU-28, subject to a number of obstacles which might reduce the value of potential increased utilisation to up to €18 billion in the shorter term and up to €134bn in the medium and longer term, depending on the scale of regulatory obstacles.

The sharing economy can be expected to reduce the cost and improve the quality of services available to consumers and create new opportunities for providers to work or increase their earnings. We do not expect problems such as road congestion or pressure on land approved for development to be exacerbated. We also do not believe income or wealth inequality will be increased, but new forms of social exclusion might result from reputation-based access controls.

We recommend new initiatives to take advantage of platform-collected data to address regulatory objectives such as limiting tax evasion at lower cost and to mitigate potential social exclusion, while preventing quantity regulation. We do not find that new policy is needed, at least for now, to address potential market power among platforms or questions over worker status. Other programmes will also need to take account of the implications of the sharing economy, e.g. those promoting the competitiveness of the European automotive manufacture industry.

PE 558.777 I - 33

AUTHORS This study has been written by Dr Andrew Lilico and Matthew Sinclair of Europe Economics, at the request of the Impact Assessment Unit of the Directorate for Impact Assessment and European Added Value, within the Directorate General for Parliamentary Research Services (DG EPRS) of the General Secretariat of the European Parliament.

RESPONSIBLE ADMINISTRATOR Pierre Goudin, European Added Value Unit To contact the Unit, please e-mail [email protected]

LINGUISTIC VERSIONS Original: EN

This document is available on the internet at: http://www.europarl.europa.eu/thinktank

Manuscript completed in January 2016. Brussels © European Union, 2016.

PE 558.777 I - 34

Table of Contents

List of tables ...... 36

List of figures ...... 36

Executive summary ...... 37

Chapter 1 – Introduction ...... 41

Chapter 2 – What is the sharing economy? ...... 42

1. Criteria for a good definition ...... 43 2. Existing definitions ...... 43 3. Our proposed definition...... 46 4. Examples of how our definition operates ...... 48

Chapter 3 – What is the Economic and Social Potential of the sharing economy in the European Union? ...... 51

1. How is the sharing economy likely to develop? ...... 52 2. Quantifying the potential sectoral scope of the sharing economy ...... 57 3. What will the impacts of the sharing economy be? ...... 63 4. Quantifying the impacts of the sharing economy ...... 79

Chapter 4 - Are there obstacles or barriers which prevent the sharing economy from reaching its full potential? ...... 85

1. Digital access and skills ...... 86 2. Physical barriers ...... 87 3. Consumer preferences ...... 88 4. Labour market obstacles ...... 90 5. Trust ...... 91 6. Tax and other policy ...... 93 7. Regulation ...... 94

Chapter 5 – What is the economic value of the barriers or obstacles? ...... 98

Chapter 6 – How effective is the existing EU regulatory framework in promoting and overseeing such emerging new business models? ...... 103

1. Related cost of non-Europe estimates ...... 104 2. Existing initiatives ...... 105 3. Remaining sharing economy issues...... 106

PE 558.777 I - 35

Chapter 7 – What additional steps should be taken at the EU level to realise its economic potential, while continuing to balance creative freedom for business with necessary regulatory protections? ...... 110

1. Improving regulations applied to sharing economy platforms ...... 111 2. Mitigating social exclusion ...... 113 3. Dealing with potential market power of sharing economy platforms ...... 116 4. Applying labour market regulation to sharing economy platforms ...... 118 5. Ensuring other initiatives reflect the growth of the sharing economy ...... 120

Bibliography ...... 122

List of tables

Table 1: COICOP categories, EU28 ...... 59 Table 2: sharing economy amenable sectors by Member State ...... 62 Table 3: Taxis and chauffeured cars, by city ...... 68 Table 4: Under-utilisation of labour ...... 80 Table 5: Under-utilisation of accommodation ...... 81 Table 6: Barrier categories ...... 99 Table 7: Obstacle effects on sharing economy potential, percentage reductions ...... 100 Table 8: Value of obstacles and barriers ...... 101

List of figures

Figure 1: Web search interest ...... 42 Figure 2: P2P vs B2C in the sharing economy ...... 54 Figure 3: sharing economy displacement in the accommodation sector ...... 56 Figure 4: sharing economy displacement in the road transport sector ...... 56 Figure 5: sharing economy displacement in the labour market ...... 57 Figure 6: Figure 4.1: Smartphone penetration, by Member State, 2013 ...... 86 Figure 7: Frequency of use of smartphone applications (social networking, camera, gaming, etc.) excluding email/messaging, 2013 ...... 87 Figure 8: Ease of Starting a Business, distance to frontier, data collection completed June 2014 ...... 96 Figure 9: Obstacles process ...... 98

PE 558.777 I - 36

Executive summary

In this report we attempt address the sharing economy: what it is; how it is likely to grow and affect consumers and providers of goods and services and its wider economic and social impacts; and how policy might need to adapt in the EU and its Member States in order to realise its economic potential while still providing necessary regulatory protections.

1. What is the sharing economy?

We define the sharing economy as the use of digital platforms or portals to reduce the scale for viable hiring transactions or viable participation in consumer hiring markets (i.e. 'sharing' in the sense of hiring an asset) and thereby reduce the extent to which assets are under-utilised. Within that definition, we include platforms with highly diverse business models allowing people and businesses to share everything from homes (e.g. AirBnB or LoveHomeSwap) to car journeys (e.g. Uber or BlaBlaCar) to drills and other consumer appliances (e.g. Peerby).

2. What is the economic and social potential of the sharing economy in the European Union?

In our view activity in the sharing economy is likely to grow in both scale and scope, extending to new markets and displacing more and less formal economic activity. Over time it will shift from being predominantly composed of peer-to-peer markets in which essentially amateur consumers share goods they own to being predominantly composed of business-to-consumer transactions with platforms offering an increasingly diverse set of services to both sides of the market.

The continuing growth of the sharing economy will affect a range of market participants:

− Consumers who will access better services at lower prices and without the need to make large purchases of expensive and often then under-utilised goods. − Providers who will see new opportunities to work or increase their earnings, though they will tend to receive less in other benefits than those in traditional employment. − Competitors who will face increased competition, sometimes in markets in which high earnings were possible as they were previously sheltered from competition.

Other potential impacts such as increased pressure on land approved for development, increased traffic congestion have been suggested, but we do not believe these problems are likely to be exacerbated except in the short term. There is also an opportunity to improve tax compliance, taking advantage of the increased use of electronic payments in sectors where cash was often the norm previously.

We also do not find the contention that the the sharing economy will increase income or wealth inequality persuasive. It seems more likely to diminish inequality and make wealth inequality, in particular, less salient. However it does seem plausible that the sharing economy could create

PE 558.777 I - 37

new forms of social exclusion if some people are unable to maintain a reasonable score in ratings systems and other systems used to protect consumers.

Overall, we expect that the potential reduction in under-utilisation of assets (including human capital) associated with the sharing economy amounts to €572bn in annual consumption across the EU28. This is, however, subject to a number of obstacles and barriers, some of which are quite fundamental.

3. Are there obstacles or barriers which prevent the sharing economy from reaching its full potential?

Barriers to the growth of the sharing economy include established features of different economies within the EU. People need to be able to use the platform applications and the smartphones on which they operate, though we expect this obstacle will decline over time. It needs to be physically practical to share the good or service in question in the market in question. This is naturally harder with bulky goods and in rural areas, though these obstacles may be overcome by new business models in some cases. In some cases, labour market obstacles such as skills mismatches may mean that demand for human capital cannot be satisfied even in markets with significant unemployment or under-employment and the sharing economy will not overcome that obstacle.

There are also obstacles that sharing economy platforms themselves can work to overcome, particularly related to consumer preferences and trust. To some extent, consumers might simply prefer ownership, but often a preference for ownership will actually reflect concerns which platforms can address to a substantial extent over time, such as over product scarcity risk or a lack of diversity in the products on offer (meaning that some consumer needs are not met by the shared stock of assets). sharing economy platforms will also face a strong incentive to overcome initial distrust through a combination of insurance, prior scrutiny of market participants and the use of ratings systems for those already participating.

Finally a range of policy choices are likely to affect the potential of the sharing economy. These include the effects of the broader policy framework, such as tax policy, which might help or hinder sharing economy providers relative to other potential suppliers of goods and services. They also include specific regulatory interventions which might go as far as directly banning sharing economy platforms or limit its growth by deterring self-employment, deterring marginal transactions or failing to provide a suitable intellectual property framework.

4. What is the economic value of the barriers or obstacles?

We expect the value of many of the barriers to decline over time. In the short-run, we expect that higher utilization of assets, facilitated by the economy, will be worth around €21bn per year (versus the €572bn obstacle-free maximum). In the medium to longer term, that figure to rise to €158bn.

PE 558.777 I - 38

That means the salience of regulatory barriers to the sharing economy can be expected to rise over time, as other obstacles such as a lack of digital access or skills and a lack of consumer trust decline.

In the short-run, we expect that high specific regulatory barriers (deterring 30 per cent of remaining transactions) would cost around €6bn a year, while low regulatory barriers (deterring 15 per cent of remaining transactions) would cost around €3bn a year. Over time, as other barriers fall away and the sharing economy grows, we expect this to rise to the point where high regulatory barriers cost €47bn a year while low regulatory barriers cost €24bn a year. That cost might rise further if other obstacles we expect to remain, such as labour market obtacles, are in fact addressed successfully.

5. How effective is the existing EU regulatory framework in promoting and overseeing such emerging new business models?

Other costs of non-Europe, established in earlier research, are likely to grow in importance with the growth of the sharing economy. Gaps in the Digital Single Market, for example, are likely to become more important to the extent they inhibit trade in or through sharing economy services.

At the same time, existing EU interventions address issues which might otherwise be expected to be significant with the growth of the sharing economy, such as the need for price transparency.

However there are some shortcomings in the existing regulatory framework. That framework does not take advantage of new possibilities to achieve regulatory objectives (such as tax collection and consumer protection) at lower cost. It also does not address some potential problems associated with the sharing economy such as the potential to create new forms of social exclusion.

6. What additional steps should be taken at the EU level to realise its economic potential, while continuingto balance creative freedom for business with necessary regulatory protections?

We finally turn to potential additional steps that might be taken at an EU level.

First, we believe there is an opportunity to increase tax compliance and otherwise accomplish regulatory objectives at a lower cost with the data collected by platforms, but that this will likely depend on removing quantity restrictions and/or assuring platforms that such restrictions will not be introduced.

Second, new measures are justified to support the rehabilitation of those excluded from platforms, including the potential establishment of community platforms for that purposes, but not through the regulation of still-evolving ratings systems.

PE 558.777 I - 39

Finally, other initiatives, such as those aiming to support the competitiveness of the automotive manufacturing sector, should take account of the potential development of sharing economy markets.

In other areas, we do not believe there is a need for new interventions. There should not be a general regulatory presumption that all sharing economy platforms will be dominant and any intervention should, for now at least, be based on an application of existing competition rules that allows for the dynamism of digital markets. Equally, labour market regulations should not be altered to specifically include sharing economy providers, who should be allowed to remain self-employed, and platforms should be allowed (and in some cases, encouraged) to develop their own means of supplying other benefits besides cash remuneration.

PE 558.777 I - 40

Chapter 1 – Introduction

At the request of the Committee on Internal Market and Consumer Protection, the Directorate on Impact Assessment and European Added Value of the European Parliament has commissioned Europe Economics to carry out an assessment of the cost of non-Europe in the sharing economy, assessing the opportunities and challenges in the sector.

The sharing economy has been the object of considerable public interest. That interest has included both excitement at the potential of the platforms to create significant economic value and concern at potential social consequences. There is not yet a consensus on the proper definition for the sector, let alone how it might affect consumers, workers and the wider economy.

This research will address those issues, but particularly the extent to which the sector might affect attempts to realise the full potential of the Single Market. The sharing economy has its own implications for the Single Market, as existing regulatory structures and tax policies might struggle to accommodate sharing economy platforms. It might also raise the salience of progress in a number of areas already identified as priorities in earlier cost of non-Europe reports.

The rest of this report aims to deepen understanding of the implications of the sharing economy for European consumers, workers and other stakeholders and establish the potential for the completion of the Single Market to add value. In doing so, it will address the following research questions posed by the Committee:

− What is the sharing economy? − What is the economic and social potential of the sharing economy in the European Union? − Are there obstacles or barriers which prevent the sharing economy from reaching its full potential? − What is the economic value of the barriers or obstacles? − How effective is the existing EU regulatory framework in promoting and overseeing such emerging new business models? − What additional steps should be taken at the EU level to realise its economic potential, while continuing to balance creative freedom for business with necessary regulatory protections?

We provide indicative quantitative estimates on those questions, where possible, particularly for the final consumption sectors in which sharing economy platforms are likely to be active; the scale of the impacts we can expect from the increased utilisation of assets through those platforms; and the value of different obstacles to the sharing economy's growth.

PE 558.777 I - 41

Chapter 2 – What is the sharing economy?

Key findings

- Interest in the sharing economy has grown quickly since 2012 but there is no settled definition of what it constitutes. - A useful definition, for the purposes of economic analysis, will not focus on an overly literal interprepation of the word 'sharing'. - Existing definitions vary but tend to focus on transactions in peer-to-peer, or often peer-to- peer, consumer markets in which an under-utilised asset is hired out. - We argue for a somewhat broader definition which includes business-to-consumer transactions, thereby focusing the analysis upon types of socio-economic transations rather than upon corporate forms.

The sharing economy has attracted considerable attention, but that attention is relatively recent. If measured by web search interest in the term 'sharing economy', then interest has mounted from 2012 onwards, as shown in Figure 1, with the greatest web search interest in Italy, Germany, the United Kingdom and the United States. This relative novelty explains the lack of a settled definition.

Figure 1: Web search interest

Source: Google Trends, 8 September 2015

PE 558.777 I - 42

1. Criteria for a good definition

For our purposes, we need a definition amenable to economic analysis and the identification of potential policy measures. In our view, a good definition of the sharing economy for this project: − Identifies a distinct and genuine phenomenon, rather than becoming unnecessarily semantic with regards to the word 'sharing'. This should enable a more robust analysis of the likely economic impacts. − Avoids artificial distinctions between economically similar activities. This could mean that policy is formed which is inappropriate for excluded, but fundamentally similar, activity. − Respects common usage, rather than requiring the use of a new term to discuss platforms that are currently commonly referred to in discussions of the sharing economy.

2. Existing definitions

An online finance glossary defines the sharing economy as an 'economic model in which individuals are able to borrow or rent assets owned by someone else' (Investopedia, n.d.). That definition could if conceived broadly include a very large range of transactions, including most or all of current rental markets in sectors like property. It would not necessarily be a new phenomenon or one which was particularly focused on the businesses typically described by those commenting on the sharing economy.

Another very broad definition is provided by a campaigning group promoting the sharing economy (The People Who Share, n.d.):

'The sharing economy is a socio-economic ecosystem built around the sharing of human and physical resources. It includes the shared creation, production, distribution, trade and consumption of goods and services by different people and organisations.

Whilst the sharing economy is currently in its infancy, known most notably as a series of services and start-ups which enable P2P exchanges through technology, this is only the beginning: in its entirety and potential it is a new and alternative socio-economic system which embeds sharing and collaboration at its heart – across all aspects of social and economic life.

The sharing economy encompasses the following aspects: swapping, exchanging, collective purchasing, collaborative consumption, shared ownership, shared value, co-operatives, co- creation, recycling, upcycling, re-distribution, trading used goods, renting, borrowing, lending, subscription based models, peer-to-peer, collaborative economy, circular economy, pay-as-you- use economy, wikinomics, peer-to-peer lending, micro financing, micro-entrepreneurship, social media, the Mesh, social enterprise, futurology, crowdfunding, crowdsourcing, cradle-to-cradle, open source, open data, user generated content (UGC).'

While this definition might suit a campaigning website, as it allows for engagement with a wide range of partners, it is too broad to be amenable to economic analysis or for use in policy

PE 558.777 I - 43

formation. The inclusion of diverse economic phenomena such as recycling, the trading of used goods and open data would undermine the coherence of any research or policy attempting to address needs or impacts of the sharing economy.

One influential recent attempt to define the sector more closely (Frenken, Meelen, Arets, & van de Glind, 2015) argued for three characteristics.

First, the authors of the definition argue it should only include consumer-to-consumer transactions, not business-to-consumer transactions. This would exclude platforms such as Spotify for music or ZipCar and Cars2Go for cars where a firm rents assets to consumers. They argue that this represents the 'product-service economy', where 'a consumer gains access to a product whilst the service provider retains ownership'. The term has been further refined by two authors from the same group, who argued that whether or not an activity should be construed as forming a part of the sharing economy turned on whether or not the asset in question would otherwise have been left idle (Meelen & Frenken, 2014). For example, they argue that 'UberX is only a form of the sharing economy if the driver would have made the trip anyway.'

Meelen & Frenken (2014) argue that definitions focused upon peer-to-peer sharing avoid the 'positive and progressive connotation' of the term sharing economy being misused by firms attempting to avoid regulatory scrutiny, but it seems unlikely that under any definition of the sharing economy it would not be subject to a potential need for regulation to protect consumers and prevent other potential market failures. We therefore do not believe that a restriction to peer-to-peer markets is particularly helpful to understanding the policy implications of the sharing economy.

The restriction to peer-to-peer transactions means that the definition is dependent on corporate form, only allowing or particularly focusing on transactions among essentially amateur consumers. This restriction can create a boundary problem. It is not clear whether mutual firms or clubs with assets held in trust for their members would be eligible and hybrid platforms may offer peer-to-peer and business-to-consumer services alongside each other. Individual providers of goods and services might also transition over time from peers, sharing assets with friends and family; to sole traders, after finding that they earn more in the sharing economy than in other employment; to a business which employs other workers, if that creates some kind of economy of scale.

While peer-to-peer transactions might have some attractions for some consumers, and those consumers might have been over-represented in the early sharing economy platforms as they find the concept particularly attractive, others might not care about the corporate structure by which a good or service is delivered (Eckhardt & Bardhi, 2015). In general, corporate form tends to be an uninformative basis for identifying economic phenomena (other than the corporate form itself). The UK mutual John Lewis Partnership and the French corporate Printemps are both department store chains in terms of their economic activity, despite their differing corporate forms.

Other sources also wish to delineate the sharing economy by corporate form, but with a restriction to transactions that are conducted within local communities, including co-operatives of various kinds and 'neighbourhood' car-sharing (Orsi, 2013). This to some extent reflects the

PE 558.777 I - 44

spirit of the definition set out for The People Who Share, but does not fit with other academic definitions or allow the potential for sharing across local and national borders.

Others still focus particularly upon non-monetised peer-to-peer transactions, suggesting that the spread of such transactions implies a profound change in the nature of capitalism.24 However, the proportion of entirely non-monetised transactions, amongst the economically- relevant activities, is currently small and not expected to grow as rapidly as the monetised transactions. Furthermore, it is important to recognise that large amounts of social activity are non-monetised in any markets-based economy.

One reason there has been a debate about whether business-to-consumer services should be included in the sharing economy appears to be that 'sharing', in the form of a non-monetised making available of one's assets or one's time to others, with a view to subsequent reciprocation by others, is seen as morally superior to trading via money-mediated transactions. But in economic theory, money-based trading is simply a mediated form of reciprocation or sharing. I could give you my economics consultancy services today in the anticipation that you would reciprocate by tomorrow giving me some commensurate amount of rhubarb pie or letting me use your lawnmower. But by exchanging the consultancy services for money I enable reciprocation and sharing to be done much more efficiently and effectively, spread over time and more people and with much better information as to my options.

Second, Frenken, Meelen, Arets, & van de Glind (2015) argue that the sharing economy should be understood to only include transactions where consumers provide temporary access to a good, not the permanent transfer of ownership of the good. This would exclude platforms such as eBay or other online marketplaces where consumers can buy and sell goods (potentially 'sharing' them over time, in the sense that multiple households will use the same asset in its lifetime). More broadly, we do not believe it is useful to see 'sharing economy' as a synonym for e-commerce. This element of the definition seems the least controversial in terms of existing attempts to define the sector closely. While digital platforms or portals25 might enable more people to trade second-hand goods, the substitution of ownership (as opposed to more frequent transfers of ownership) is widely seen as a defining quality of the sharing economy.

Third, Frenken, Meelen, Arets, & van de Glind (2015) argue it should only include transactions regarding physical assets. The restriction to transactions regarding physical assets would exclude the many websites allowing people to provide services, such as TaskRabbit. The

24 See, for example, Mason (2015). 25 For our purposes here, a 'digital platform' is a digital technology or format that allows a range of suppliers a platform from which to present themselves to potential consumers and, conversely, for potential consumers to access a range of potential suppliers. Two well-known example from other contexts are (i) the Sky Digital platform, which allows a range of television stations to present themselves to viewers; and (ii) Facebook, which is commonly referred to as a digital platform for social media users. A sharing economy digital platform is a digital technology allowing those that have assets they want to share to present themselves to those seeking assets to share and vice versa. A 'digital portal' is a website designed to bring information together from a range of sources in a standard format. In the case of the sharing economy, a digital portal is a digital technology that brings together other information about assets that users want to share and those seeking assets to share in a way that allows the two sides of the sharing transaction more easily to discover and deal with each other. To avoid clumsy repetition of terms and because platforms are the main form of the sharing economy and the form that creates the most important issues (positive and negative), in what follows we shall normally refer to 'platforms', but portals should be considered as implicitly included where relevant.

PE 558.777 I - 45

authors argue that people cannot go 'unused', and the value of time away from work as leisure time might mean that under- or unemployment has more potential to be attractive in the labour market, but people can be employed, under-employed or entirely unemployed, just like goods.

Excluding labour seems to introduce an undesirable artificial limitation on the definition. Another way to understand sites like TaskRabbit is that they allow people to share their human capital (e.g. their expertise in assembling flat pack furniture), with those who lack that human capital (e.g. those intimidated by the process of assembling flat pack furniture). It is also useful to recognise that all sharing economy activities involve some combination of the use of labour (by the provider) and/or the avoidance of labour (by the consumer).

A restriction to physical assets could also exclude the sharing of intellectual property. This could mean that, for example, a song shared on a CD would count as a part of the 'sharing economy' but a song shared as a digital file would not. That artificial distinction might become increasingly problematic over time if an increased salience of design features and technologies like 3D printing meant that the intellectual property component increased as a share in the value of assets previously best-understood simply as physical goods.

Others proposing a definition for the sector have not included such a restriction to physical assets. Botsman (2013), for example, defines the sharing economy as an 'economic model based on sharing underutilized assets from spaces to skills to stuff for monetary or non-monetary benefits.' She also does not exclude business-to-consumer transactions, but does say that the sharing economy is 'largely based on P2P marketplaces.' And Wosskow (2014) defines it as allowing people to 'share property, resources, time and skills across online platforms.'

We believe that these existing sharing economy definitions are a helpful starting point but are either too broad, and do not represent a distinct economic phenomenon amenable to analysis, or are too narrow, attempting to delineate among different business models delivering very similar goods and services from the perspective of the consumer, in quite similar ways from an economic perspective. We therefore propose to build on these definitions and provide a new definition of the sharing economy as an economic phenomenon.

3. Our proposed definition

We define the sharing economy as:

We therefore define the sharing economy by two elements:

First, the sharing economy is an opportunity to reduce the extent to which assets are utilised less than they could be by employing a rental model. Often a single transaction, e.g. buying a car, is replaced by a stream of smaller rental transactions, e.g. renting a car each time it is needed for a journey. Under-utilisation might be reduced because someone is able to hire out

PE 558.777 I - 46

their assets which would otherwise go unused, or because they are able to hire someone else's assets and therefore do not have to purchase functionally-equivalent assets themselves. Second, it is made possible by advances in information and communication technology which have reduced transaction costs and increased the extent to which sharing, which under any definition is not new, can become increasingly important as more people can enter those markets (even if the scale of the economic opportunity for them is small) and it can satisfy the needs of a broader range of customers (even those whose business only represents a small opportunity). Taking advantage of that opportunity to extend rental markets constitutes the sharing economy.

To some extent, our definition of the sharing economy reflects what others have described as broader categories including the sharing economy, such as the Collaborative Economy (Botsman, 2013), or includes sectors which others have defined as similar to but not a part of the sharing economy, those transactions not based on a peer-to-peer model have been called the Product-Service Economy (Frenken, Meelen, Arets, & van de Glind, 2015). While none of these terms are necessarily illegitimate, the sharing economy is by some margin the most commonly used and a reasonable descriptor which reflects common usage for the entire economic phenomenon we will study, rather than a part of it.

We have focused our definition on consumer markets (i.e. peer-to-peer or business-to-consumer markets), as opposed to pure business-to-business markets. While platforms similar to those used in the sharing economy might gain scale connecting businesses (particularly small firms), and the corporate marketplace has been discussed as a potential growth sector for new and existing platforms (Slagen, 2014), this seems like a separate phenomenon that might have quite different economic impacts and policy implications. Equally, the sharing economy can be seen as the extension of trends which have already been taking place within the corporate world for some time, with specific firms specialising in owning assets and providing them to others as a service (e.g. aircraft leasing). Digital technologies allow the sharing economy to be created for consumers, when previously this extension of the rental model required the scale and organisational capacity of a corporation. We will, however, address the potential for consumers other than individual households to use sharing economy platforms (e.g. local governments).

In our definition, the sharing economy could include activity in platforms providing access to the following goods and services:

− accommodation; − transportation; − consumer durables; − labour and human capital; and − intellectual property.

We might also include finance, but finance falls outside the scope of our terms of reference in this study. The role of the sharing economy in finance might be quite different and the regulatory considerations are unique and it is therefore set aside here for consideration in other research.

PE 558.777 I - 47

4. Examples of how our definition operates

4.1. Accommodation

There are a number of platforms which allow people to rent out properties or parts of properties. The most famous is AirBnB, but there are a number of competitors including: HomeAway, HouseTrip, 9Flats, Wimdu, Onefinestay, Roomerama, Sleepout, Love Home Swap and Holidaylettings.

If someone had a property which was wholly or partially unoccupied they might previously have allowed friends and family to stay in a spare room or a holiday home, an option which already existed as a substitute for hotels, bed and breakfasts. However the practice was limited by the high cost of matching those with accommodation free to those in need of accommodation, particularly beyond borders and establishing trust on both sides of transactions between strangers.

Information and communication technology, in the form of the platforms mentioned above, and the hardware that makes them possible, has significantly reduced the broadly-defined marketing cost of offering private rental occupation. Digital platforms and portals have thereby extended the potential of that rental model to short-term rentals which might only be a marginal use for a property principally used as a main residence by the owners. It might thereby reduce the degree to which assets are under-utilised by allowing people to make better use of otherwise unoccupied property.

We observe that there are some potentially relevant differences between the services offered by the platforms we have listed above. Some offer holiday rental services. Others specifically focus upon peer-to-peer rentals or describe themselves as an 'online marketplace'.

4.2. Transportation

There are two types of transportation service which fall under our broader definition of the sharing economy. First, the hiring of the assets themselves: hiring a car, bike or other mode of transport. Firms in this segment include: ZipCar, Car2Go and Autolib' for cars and bike share schemes often organised by city with municipal involvement, such as the Velib' scheme in Paris. Second, the hiring of an asset mixed with labour and human capital: hiring a car or other vehicle and someone to drive it. The most prominent firms in this segment are Uber and Lyft.

Rental services for cars and other vehicles again existed before the advent of the sharing economy (as did individuals sharing their cars with friends and family in need of one for an occasional journey). Information and communication technology has reduced the scale required in terms of the individual transaction: people might rent cars for an hour as technological change has simplified the process dramatically. It has also reduced the scale required of the market opportunity needed for it to be worthwhile to enter the market.

The relative simplicity and low cost of becoming a driver for a service like Uber reduces barriers to entry. This can further extend the rental market and thereby reduce the extent to which assets

PE 558.777 I - 48

are under-utilised both by allowing people to make more use of their cars but also by allowing others to avoid buying a car which would otherwise be under-utilised.

4.3. Consumer durables

Peerby, NeighborGoods and others allow people to share other consumer goods such as drills, trailers, barbecues, suitcases and garden scissors.

These goods may well have been shared in the past among friends, family and neighbours. Information and communication technology has again reduced the cost of marketing their availability, to the point where it is worthwhile to offer them to be borrowed to strangers for free or rented for a fee, at least within a reasonable geographical area.

4.4. Labour and human capital

TaskRabbit, Skillshare, 99Designs, Kaggle, Shareyourmeal and others allow people to share their labour and human capital. There is also a labour element in a number of other services. In some cases, that labour element is quite large, e.g. DogVacay or UberX. In others, it is relatively modest but still might be a substantial part of why consumers chose these services over ownership. For example, renting a car from ZipCar might be preferable to owning a car for some consumers in part because they felt that the challenges of ownership – arranging maintenance, paying taxes and even identifying a suitable car in the first place – was a job which they did not want to do. This still represents a potentially-significant division of labour.

It was possible to hire people for all of these tasks before the advent of the sharing economy, however again the rental model is extended with more tasks being undertaken with short-term contracts, including some work which might previously have taken place outside the labour market, and that reduces the extent to which labour or human capital is under-utilised.

4.5. Intellectual property

Spotify, Apple Music, Pandora, Tidal, Rhapsody, Google Play and others (YouTube might be counted in this category), and similar services for video often offered as part of a broader package of services by broadcasters, allow users to rent intellectual property in digital formats temporarily, rather than purchasing them permanently.

Before online distribution became feasible, customers could buy, rent or steal content from those who owned the right to distribute that intellectual property. Rental services were limited by the size of the transaction. While movies were rented on videotapes and then DVDs, such services were less common with music where being able to listen to a CD for a short period of time was not valuable enough to justify the transaction costs of a short-term rental. Radio allowed people to listen to songs one at a time but removed the listener's ability to choose what was played when.

PE 558.777 I - 49

Of the three potential models for the distribution of music, at first peer-to-peer services increased the ease of pursuing the illicit option, with services like Napster. iTunes and similar services then extended the purchase option. Finally, Spotify and its competitors extended the rental model, establishing a means for consumers to access music on a track-by-track basis and thereby extending the rental model by reducing the viable size for a given intellectual property rental (to a single track for a single play).

PE 558.777 I - 50

Chapter 3 – What is the Economic and Social Potential of the sharing economy in the European Union?

Key findings

- The nature of the sharing economy is likely to change over time as its scope and scale grow, and as it matures - Consumers will tend to benefit from lower prices and increased quality of services, along with a reduction in the lumpiness of their consumption and the ability to satisfy more diverse preferences over time. - Providers may enjoy new economic opportunities, but may not enjoy access to the other benefits associated with traditional employment. - Competing providers outside the sharing economy may face increased competition in service markets, but the sharing economy will generally not increase the pressure on scarce resources such as land for development or road space. - Manufacturers may need to adapt to a market in which fewer, but higher value, goods are consumed. Other policy priorities will also be affected, with a reduction in income and wealth inequality but the potential to create new forms of social exclusion. The use of electronic payments and digital platforms makes a significant improvement in tax compliance possible. - The obstacle-free theoretical maximum potential reduction in under-utilisation associated with the sharing economy amounts to €572bn, though that is subject to a number of (in some cases quite fundamental) barriers discussed in Chapter 4.

The sharing economy has inspired excitement at the scale of the opportunity for the many people engaged in it and its potential wider benefits and concern at the potential impact on established sectors and working practices. On the one hand, a supportive campaign describes the sharing economy's potential in glowing terms (The People Who Share, n.d.):

'Our vision of the future is a thriving sharing economy where the need to own is transformed. Everyone is a supplier of goods, services and experiences. Where people share skills, time, resources, knowledge, responsibility, opportunities, ideas, goods, services and stuff. It's a world in which our collective capability meets our collective needs and we collaborate to enhance each other's lives, protect our planet and create wealth from which everybody benefits.'

On the other hand, a sceptical commentator describes the sharing economy as a dystopia (Reich, 2015):

'How would you like to live in an economy where robots do everything that can be predictably programmed in advance, and almost all profits go to the robots' owners?

PE 558.777 I - 51

Meanwhile, human beings do the work that's unpredictable – odd jobs, on-call projects, fetching and fixing, driving and delivering, tiny tasks needed at any and all hours – and patch together barely enough to live on.'

The gap between the two is in part a difference over what the sharing economy is and what it is likely to become and in part a difference over its likely wider effects.

In this chapter we will first set out very high-level expectations for how the sharing economy is likely to develop. Then we will attempt to quantify the potential scale of the sharing economy in terms of the sectors in which it is likely to compete effectively with other models for delivering goods and services. Then we will set out the likely impacts of that development and then we will attempt to quantify the likely impacts.

1. How is the sharing economy likely to develop? We defined the sharing economy in Chapter 2 and gave examples of existing sharing economy platforms. The potential of the sharing economy will depend on how it develops, with the growth or contraction of existing platforms and the creation of new platforms or changes in existing platforms, potentially extending the sharing economy to new sectors.

However we can describe at a high level how the sharing economy might develop based on the definition and the examples given above. This development will then be subject to the obstacles described later in this report:

1. The sharing economy is likely to extend to other markets. 2. Peer-to-peer transactions are likely to decline as proportion of the sharing economy, with business-to-consumer transactions accounting for a larger share. 3. sharing economy platforms are likely to become more integrated with other platforms and other digital services. 4. sharing economy platforms will displace more and less formal activity.

1.1. The sharing economy is likely to extend to new markets

While some new platforms will represent substitutes for existing platforms, others are likely to cover sectors, geographical settings and consumer groups where the reach of the sharing economy is currently limited.

It is obviously difficult to speculate about the markets to which it might extend, as that will reflect the business models of new entrants, but those sectors where the potential to be affected by the sharing economy is greatest might include:

− Sectors in which there are valuable goods and services still under-used (e.g. smaller domestic appliances) and in which sharing economy platforms can materially reduce existing barriers to entry, particularly those relating to marketing and other transaction costs.

PE 558.777 I - 52

− Sectors which are likely to grow as a share of the economy as a whole, thanks to high income elasticities, e.g. education and health. These sectors might be affected directly by the sharing economy or indirectly if savings in other areas directly affected are spent in those sectors. − Markets in which sharing economy platforms are not currently able to attain critical mass, but where they might expand with growing scale and/or technical changes, e.g. from urban to suburban or even rural areas.

1.2. Peer-to-peer transactions are likely to decline as proportion of the sharing economy

Many early applications of the sharing economy have been based on peer-to-peer markets. Indeed some definitions of the sharing economy define it entirely as a peer-to-peer phenomenon. In many markets, however, the peer-to-peer element seems likely to be a feature of a transition to the sharing economy.

Initially most people might own an asset, as the transaction costs involved in renting them are high and rentals are therefore reserved for niche uses (e.g. when travelling, or for expensive assets rarely used like a van for moving). sharing economy platforms reduce those transaction costs and therefore people are able to share the assets they already own. Given that the costs of purchasing those assets have already been paid, P2P providers are therefore competitive. Over time, however, the declining attraction of ownership means that fewer people replace the assets they own as those assets reach the end of their natural life. P2P rentals then become less attractive as more customers chase a shrinking pool of assets, resulting in greater product scarcity risk and greater costs in matching providers and consumers (e.g. in transporting an asset to its potential user). B2C rentals then increase their market share.

This intuition is illustrated in Figure 2, which shows a simple model in which ownership declines by one per cent of the population in each time period – t – (as renting through the sharing economy becomes more attractive) and the share of the renting market accounted for by P2P reflects the share of ownership in the market as a whole (i.e. if ten per cent of the market owns, P2P will account for ten per cent of rental transactions). This model is not intended to be realistic and clearly a far more elaborate analysis of consumer choice would be needed to model these trends in detail, but is helpful to illustrate the intuition described above, which reflects the thinking of many involved in the sharing economy.

The trend in that model is that at first the P2P share grows strongly, as the rental market is growing and it is the favoured means of satisfying that growing rental demand, but the P2P share of the market as a whole peaks at around 25 per cent (when ownership is around 50 per cent) and then begins to decline. Eventually the market is dominated by B2C rental. The increasing scale of P2P transactions now might represent the start of such a process.

PE 558.777 I - 53

Figure 2: P2P vs B2C in the sharing economy

It is important to note two things about this analysis, though:

First, B2C provision does not necessarily imply corporate provision. B2C providers may be individuals, e.g. drivers using a platform like Uber; co-operatives such as car clubs; or corporates such as car sharing companies. The intuition is rather that people will increasingly engage in these markets not as peers but that the market will specialise with some owning assets in order to rent them, rather than renting out assets they own primarily for their own purposes.

Second, it may not apply in all markets. In some markets, people might have a preference for ownership (e.g. in property) and use the income from sharing economy platforms in order to be able to afford ownership (e.g. it might make holiday homes more affordable). In that case, a substantial P2P market might remain viable over time. In other cases, providing a service yourself might have a cost advantage, which means it remains more viable to share that service with peers than to outsource provision.

Long-distance journeys by road, for example, might be more affordable if someone does not have to pay a driver who would otherwise not be making the same journey. That might mean there is an ongoing market for P2P services such as BlaBlaCar, if they can overcome other economies made possible by B2C provision, e.g. coaches and other large vehicles which provide economies of scale.

PE 558.777 I - 54

1.3. sharing economy platforms are likely to become more integrated with other platforms and other digital services

As the sharing economy matures, platforms might become less free-standing. As well as increasing their engagement with other existing firms (such as insurers) they are likely to increase their interactions with one another and with other digital services.

There are two levels on which this might occur.

First, services provided to sharing economy providers, allowing them to overcome obstacles to its growth. Zhuo (2015) cites 'Zen99, which helps 1099 workers handle their taxes, and Checkr, which provides automated background checks' as examples of this trend. These services might be heavily dependent on the policy environment, for example whether data protection regulations or the organisation of police background checks allows for automation.

Second, matching consumer needs across sharing economy platforms and other digital services. Zhuo (2015) describes a model in which 'people might use OpenTable to not only get a dinner reservation, but to also order a Lyft, buy flowers and book an overnight stay.'

This might be a means by which new firms are able to enter a market in which an existing platform exists, either partnering with an existing competitor or establishing their own platform in order to build critical mass more quickly using established online customer relationships.

1.4. sharing economy platforms will displace more and less formal activity

In discussions about the sharing economy, it is common to assume that sharing economy services only displace more formal or more regulated activity, e.g. that a rental on Airbnb will displace a potential stay in a hotel. There are actually a number of possibilities:

− The activity would otherwise not have taken place, e.g. people might take more trips with new options for where they might stay. − The activity would otherwise have taken place in a more formal sector, e.g. a hotel. − The activity would otherwise have taken place in a less formal sector, either someone might have stayed with friends or family or they might have found a place with a stranger through less formal channels.

The potential for substitution in the accommodation sector is shown in Figure 3.

PE 558.777 I - 55

Figure 3: sharing economy displacement in the accommodation sector

There is a similar set of potential substitutions in the road transport sector.

Figure 4: sharing economy displacement in the road transport sector

In the labour market, such substitution is possible but it is also important to delineate between existing employment and self-employment, where the associated conditions for providers are more similar to those generally offered on sharing economy platforms.

PE 558.777 I - 56

Figure 5: sharing economy displacement in the labour market

2. Quantifying the potential sectoral scope of the sharing economy

2.1. Earlier estimates

There have been several attempts to estimate the potential size of the sharing economy in broad terms, with estimates for global revenues ranging from $110bn to $530bn (PWC, 2014). PWC research recently estimated that global revenues from sharing economy platforms in five sharing economy sectors – peer-to-peer finance (excluded from this study), online staffing, peer- to-peer accommodation, car sharing and music video streaming – have the potential to increase from around $15bn now to $335bn by 2025. Crucially, this analysis is based upon extrapolating the future size of the sharing economy from its present size and growth. The researchers reviewed existing revenue data and forecasts where available (normally for less than five years) and then calibrated that to an S-curve, where an industry goes through stages from a slow- growing niche, to growing rapidly and then normalising to growing more slowly as it matures and then eventually declining (or being reborn with some fresh innovation).

While it provides a sense that some of the firms and sectors involved have the potential to grow considerably, and is therefore of interest to those attempting to forecast potential growth for those firms, that does not estimate the potential of the sharing economy as a whole. The sharing economy's future development might be dominated by new potential business models in existing sectors expanding out of niches in a continuous process over time, or an expansion into new sectors entirely (changing the implications for policy). The approach is therefore poorly suited to our purposes.

Some commentators estimating the potential of individual firms have focused instead on understanding the markets in which sharing economy platforms might displace other modes of

PE 558.777 I - 57

distribution. One criticism of the valuation of Uber argued that it would require the firm to account for an unrealistic share of the global taxi market, estimated at around $100bn in size (Damodaran, 2014). In response, an investor in the firm argued that the maximum market share envisaged was unrealistic and that it was wrong to see the firm as competing within the existing taxi market, arguing instead that it would expand the car for hire market to take in journeys that would not otherwise have taken place and displace the ownership of private cars or their use (Gurley, 2014). This approach, focusing on establishing the sectors of the economy in which sharing economy platforms are likely to permit an extension of the rental model, seems to fit best with the definition that we have adopted for the sharing economy.

Our approach, at this stage, is therefore to identify those sectors where either:

− households obtain an imputed value from assets they might no longer need to own, substituting with provision from sharing economy platforms; or − households purchase goods or services from markets where sharing economy platforms already compete with other provision.

2.2. Identifying sectors in which sharing economy platforms might compete

Our first step is to identify different sectors where sharing economy platforms exist or might develop. Given that we define the sharing economy as a consumer market phenomenon, we use final consumption sectors based on the COICOP26 3-digit codes used by Eurostat. The sectors are set out in Table 1 below. They are first categorised as representing either for perishables (in this case, defined by markets in which consumers could not viably trade goods they had purchased), goods or services. Then final consumption per household across the EU28 is given, to the nearest €100, the most relevant asset for that sector's potential interaction with the sharing economy (for many goods sectors, this is the sector itself) and examples of existing sharing economy provision if platforms already exist (with a note where those platforms exist in the United States but not yet in Europe).

Of average per capita consumption of €14,800 a year in the EU28, we can establish the following categories:

− €5,000 relates to spending in categories entirely or largely related to perishables, where sharing economy penetration is therefore likely to be limited; − €1,200 relates to spending in categories which seem to be more amenable to other digital business models, particularly platforms allowing the trading of second-hand goods, and the sharing economy therefore seems likely to remain a niche. − €1,200 relates to spending on financial services, excluded from this project, or the activities of non-profit organisations serving households rather than consumer markets. − €600 relates to services tied to major facilities such as museums or hospitals.

26 Classification of Individual Consumption According to Purpose – a nomenclature developed by the United Nations Statistics Division for individual consumption expenditures incurred by households, non- profit institutions serving households and general government according to their purpose.

PE 558.777 I - 58

− €6,800 relates to sectors where there is already significant sharing economy activity or where analogous sharing economy platforms seem plausible, these are the markets we believe are amenable to sharing economy business models.

Around half of this consumption (46 per cent) is in markets amenable to sharing economy business models. It should be noted that this does not indicate a limit on the value of the sharing economy. sharing economy platforms might increase GDP, leading to increases in the consumption of certain goods and services, and they might involve the provision of services not provided by traditional provision. It does, however, provide both an indication of the potential scope of existing sectors that appear likely to be affected by the sharing economy (around half of consumption) and a sense of the expenditure involved in purchasing assets which might go under-utilised (and therefore the potential for savings with reductions in the under-utilisation of those assets).

Table 1: COICOP categories, EU28

Classification of Individual Consumption According € per Principal Existing sharing economy Type to Purpose (COICOP), 3 person relevant asset provision? digit aggregates Food and non-alcoholic Perishables 1,900 - beverages Alcoholic beverages, Perishables 500 - tobacco and narcotics Most platforms focus on the Clothing and Clothing and footwear Goods 800 second-hand trade, rather footwear than sharing as defined here. Actual rentals for housing Services 700 Accommodation Yes, e.g. Airbnb. Imputed rentals for Goods 1,800 Accommodation Yes, e.g. Airbnb. housing Maintenance and repair of Goods 100 Accommodation Yes, e.g. Airbnb. the dwelling Water supply and miscellaneous services Perishables 300 -

relating to the dwelling Electricity, gas and other Perishables 700 - fuels Sites where accommodation Furniture and furnishings, is shared will generally involve carpets and other floor Goods 300 Furniture sharing fully-furnished coverings accommodation. None, but this seems analogous to a combination of Household Household textiles Goods 100 existing laundry services (e.g. textiles Laundrapp) and durables sharing services (e.g. Peerby). Household Household appliances Goods 100 Yes, e.g. Peerby. appliances

PE 558.777 I - 59

Classification of Individual Consumption According € per Principal Existing sharing economy Type to Purpose (COICOP), 3 person relevant asset provision? digit aggregates Most platforms focus on the Glassware, second-hand trade, rather Glassware, tableware and tableware and than sharing as defined here. Goods 100 household utensils household There is established rental utensils provision for glassware for occasional use. Tools and Tools and equipment for equipment for Goods 100 Yes, e.g. Peerby. house and garden house and garden Tools and Goods and services for equipment for routine household Goods 200 house and Yes, e.g. TaskRabbit. maintenance garden, human capital This category largely relates Therapeutic to perishables such as Medical products, Perishable/ 200 appliances and pharmaceuticals, but also appliances and equipment Goods equipment appliances that might be shared. Yes, e.g. Pager (US). Individual consumption may make up a small share of overall Medical spending on out-patient Out-patient services Services 200 equipment, services, but this does not human capital preclude a sharing economy role in private or public provision. Medical No, this is generally tied to Hospital services Services 100 equipment, facilities only available in human capital larger facilities. Purchase of vehicles Goods 500 Vehicles Yes, e.g. car and bike sharing. This category largely relates Operation of personal Perishable/ Vehicles, human to perishable motor fuel, but 1,000 transport equipment Goods capital also appliances that might be shared. Vehicles, human Yes, e.g. car and bike sharing Transport services Services 400 capital and Uber. No, second-hand trading is more common. This category is more relevant due to the Communications Services 400 Telephones use of smartphones as the principal means by which users access sharing economy services.

PE 558.777 I - 60

Classification of Individual Consumption According € per Principal Existing sharing economy Type to Purpose (COICOP), 3 person relevant asset provision? digit aggregates Platforms have existed for this purpose – e.g. Snapgoods (US) Audio-visual, photographic – although it does not seem and information Goods 200 Cameras to be an active market at processing equipment present. Other forms like second-hand trading are more common. Other major Other major durables for durables for Goods 0 Yes, e.g. Peerby. recreation and culture recreation and culture Accommodation, other Other recreational items recreational Yes, e.g. Peerby or DogVacay and equipment, gardens Goods 300 items and (US). and pets equipment, gardens and pets Cultural No, this is largely based on Recreational and cultural facilities, such as Services 500 public institutions such as services libraries and galleries or museums. galleries Newspapers, books and Newspapers and Goods 200 Yes, e.g. Blendle. stationery books Yes, e.g. Airbnb. Packaging with flights not offered, but Package holidays Services 100 Accommodation this does not seem a fundamental difference for our purposes here. Learning Education Services 200 materials, Yes, e.g. Khan Academy. human capital Yes, e.g. Kitchensurfing, Restaurants, although it seems plausible Catering services Services 1000 cafés, canteens, this will remain a niche human capital service. Accommodation, Accommodation services Services 200 Yes, e.g. Airbnb. human capital Limited, but seems analogous Electric to other services for which appliances for Personal care Services 300 sharing economy providers personal care, have been able to such as human capital mobility services. Jewellery, clocks, watches None. Second hand trading Personal effects n.e.c. Goods 100 and other more common. personal effects

PE 558.777 I - 61

Classification of Individual Consumption According € per Principal Existing sharing economy Type to Purpose (COICOP), 3 person relevant asset provision? digit aggregates Assets of non- profit Social protection Services 200 institutions Not relevant to this project. serving households Insurance Services 400 Financial assets Not relevant to this project. Financial services n.e.c. Services 400 Financial assets Not relevant to this project. Assets of non- profit Other services n.e.c. Services 200 institutions Not relevant to this project. serving households Total 14,800

Source: Eurostat, Final consumption expenditure of households by consumption purpose - COICOP 3 digit - aggregates at current prices [nama_co3_c]

The breakdown by Member State is shown in Table 2. The portion amenable to the sharing economy is reasonably stable at around 45 to 50 per cent of consumption, but is lower in some new Member States, where perishables account for a considerably higher share of overall consumption (to be expected with lower per capita incomes).

Table 2: sharing economy amenable sectors by Member State

Other Financial sharing business services Requires sharing Perisha- economy models or not major economy Total bles amenable more consumer facilities amenable portion suitable markets EU28 5,000 1,200 1,200 600 6,800 14,800 46% Belgium 6,000 1,400 1,700 1,000 7,300 17,400 42% Bulgaria 1,700 200 100 100 1,100 3,200 34% (2011) Czech Republic 3,200 300 200 300 3,200 7,200 44% Denmark 6,800 1,600 2,100 900 9,800 21,200 46% Germany 5,800 1,500 1,600 800 8,100 17,800 46% Estonia 3,200 600 200 200 2,500 6,700 37% Ireland 5,200 1,000 800 1,000 8,500 16,500 52% Greece (2011) 5,700 800 500 500 6,600 14,100 47% Spain 4,700 900 700 500 6,800 13,600 50% France 6,200 1,200 1,400 600 8,100 17,500 46% Italy 5,400 1,500 900 600 7,200 15,600 46% Cyprus 4,800 1,000 600 900 7,500 14,800 51% Latvia 3,300 400 100 200 2,600 6,600 39%

PE 558.777 I - 62

Other Financial sharing business services Requires sharing Perisha- economy models or not major economy Total bles amenable more consumer facilities amenable portion suitable markets Lithuania 3,900 600 200 100 2,200 7,000 31% Luxembourg 11,000 2,000 2,300 600 14,600 30,500 48% Hungary 2,600 200 400 200 2,000 5,400 37% Malta 4,000 800 600 700 5,400 11,500 47% Netherlands 5,100 1,300 1,900 400 7,300 16,000 46% Austria 6,400 1,800 1,300 1,000 9,600 20,100 48% Poland 2,800 500 600 200 2,000 6,100 33% Portugal 4,000 900 700 300 4,300 10,200 42% Romania 2,200 100 0 100 1,300 3,700 35% (2010) Slovenia 4,700 800 700 400 3,900 10,500 37% Slovakia 3,400 400 300 200 2,900 7,200 40% Finland 5,900 1,400 1,300 1,000 9,500 19,100 50% Sweden 6,700 1,500 1,600 900 9,700 20,400 48% United 5,500 1,800 1,300 800 9,600 19,000 51% Kingdom

Source: Eurostat, Final consumption expenditure of households by consumption purpose - COICOP 3 digit - aggregates at current prices [nama_co3_c]

The sharing economy will have a range of impacts in those sectors amenable to its growth, and in the wider economy, which we will consider in the next section, depending on its ability to overcome obstacles discussed in Chapter 4.

3. What will the impacts of the sharing economy be?

3.1. Consumers

Consumers in these markets are generally individuals and private households, but businesses and public sector bodies can also make use of sharing economy platforms. Local authorities often use vehicles in their own fleets relatively infrequently, for example, and therefore incur relatively high costs which might be reduced using sharing services (Vanhee, 2011, p. 10)

3.1.1. Prices

Consumers can be expected to engage with sharing economy platforms to the extent that they are better off for doing so, paying a lower quality-adjusted price for the good than if they bought it by some other means. Prices can be lower for consumers using sharing economy platforms, for three reasons:

PE 558.777 I - 63

− Increased utilisation. To the extent that shared assets are used more, and the fixed costs of owning them can therefore be spread over more use, prices for a given volume might fall. − Increased supply. To the extent that prices were previously high as a result of some obstacle which prevented the expansion of traditional supply, e.g. restrictions in France on road transport services like long-distance coaches, then prices might fall with increased supply using sharing economy platforms, e.g. the rides available using the BlaBlaCar. − Lower costs. To the extent that the transaction costs constituted a significant part of the overall cost of delivering a given good, e.g. music, then prices might fall with sharing economy platforms reducing those costs and sufficient competition between them that those savings are passed through to consumers.

Around 76 per cent of respondents in a US survey reported that they agreed or strongly agreed that sharing saves money (PolicyInteractive, 2014). There is some evidence this is true in practice:

Average monthly transportation costs were found by different studies to have decreased by between $154 and $435 for US car sharing members and CA$392 to CA$492 for Canadian car sharing members (Shaheen, Cohen, & Chung, 2009).

In a survey of low-income Los Angeles communities, Uber research found that the platform provided a reduction in average journey costs of more than half (Smart, et al., 2015).

Many peer-to-peer services offer assets themselves at a zero cost, with BlaBlaCar for example only asking the person being carried to make a contribution to the driver's costs (e.g. motor fuel). The idea is simply to split the savings from sharing the largely fixed costs associated with the journey.

3.1.2. Quality

A key stated aspiration of sharing economy platforms is that the consumer experience should be better than that for the nearest traditional transaction. Indeed, in some cases that is claimed to be an automatic consequence of the 'sharing' — e.g. that someone sharing their home will offer more intimate hospitality than a traditional hotel. It is worth noting, however, that this might be mainly an 'early adopter' effect arising from some combination of those coming on board seeking to establish a good reputation and enjoying the novelty of the transaction, with the effect declining over time (Eckhardt & Bardhi, 2015).

Even leaving aside an intrinsic value to sharing though, there are three channels by which quality might be improved:

− Enhanced transparency, through public ratings systems. These provide an incentive for improvement in the quality of the service, as poor service might result in a provider being removed from the platform and good service might mean they enjoy more business or can charge higher rates (Airbnb encourages newer providers, who have not established a track record, to charge lower rates). This naturally depends on the ratings

PE 558.777 I - 64

being reliable. There is concern that Airbnb reviews might be biased by customers being concerned that leaving a bad review might lead to them being seen as a difficult customer, limiting their future rental opportunities (Mulshine, 2015), for example. There is a robust incentive for platforms to overcome these challenges and maximise that reliability, however. − Increased competition leading to improvements in new and existing providers. To the extent existing providers face enhanced competition as a result of the growth of the sharing economy, they might respond by increasing service standards. There is some evidence that this effect is already being felt, with the growth of Uber in New York and Chicago coinciding with a decline in complaints about conventional taxis (Wallsten, 2015). Features such as a robust dispute resolution might become an advantage in a competitive market and weaknesses in those systems would be a reputational risk. − New innovations might be reflected more quickly in the capital stock. As the frequency at which consumers choose which asset to consume, there is an increased incentive to reflect technological improvements more quickly. This trend has been observed in the software sector with the development of 'software as a service' business models, where it creates an incentive for greater investment in product development, higher software quality, greater profits for software publishers and higher consumer and social welfare (Choudhary, 2007). The same can be seen in goods markets where rental cars are generally younger. The British Vehicle Rental and Leasing Association claims that the 'average age of cars on European roads is eight years old', but the 'average rental car is less than eight months old' (BVRLA, n.d.).

To the extent that these changes mean the sharing economy is associated with increased quality, the question of whether or not providers are subject to requirements for professional qualifications and other regulatory controls might become less salient. It might be possible that such a shift if emphasis from regulatory supervision to market discipline would increase average quality, but lead to greater variance as lower-quality providers repeatedly enter these markets and are then unable to compete in a relatively well-informed market. This point will be considered further in Chapter 6.

3.1.3. Reduced lumpiness

The smaller transactions which can be economical when using sharing economy platforms may further increase consumer welfare by either:

− allowing transactions to take place which might otherwise not have justified fixed transaction costs; or − reducing the extent to which consumers pay for unwanted goods and services.

The impact of the reduction in lumpiness can be seen most clearly in car sharing services. First, there is an increase in access. 60 per cent of households joining car sharing households in North America were previously carless (Martin & Shaheen, 2011).

An increase in access might manifest itself as an increase in consumption in some markets. There might be a wide range of music that a consumer does not expect to enjoy enough to

PE 558.777 I - 65

justify the fixed cost of purchasing a CD or a song, for example, but where they can justify the extremely low (often zero) incremental cost of renting it for a single play.

Second, there may be a reduction in overall costs. The reduction in costs noted for users of car sharing services may partly reflect reduced lumpiness, as well as lower prices. There was an observed increase in walking and the use of public transport (Shaheen, Cohen, & Chung, 2009), which might reflect that (when adjusting for other benefits) the cost of those modes of transport was lower. Consumers had previously been forced to pay the substantial cost of buying and maintaining a car in order to access one, after which that cost was sunk and not reflected in their decisions over the transport mode for each subsequent journey.

For more valuable assets, such as cars, reduced lumpiness, a smoothed consumption profile, may also mean a straight-forward reduction in financing costs as consumers no longer need to borrow in order to buy an asset, or consume savings which might otherwise be invested in assets generating a financial return.

Reduced lumpiness might also mean that the range of and scope for regulatory under-cutting increases. Where there are material fixed costs anyway, relatively small additional fixed costs from establishing proper regulatory credentials and compliance may add only modestly to total transaction costs, meaning the incentives to undercut regulation by non-compliance are limited. But if the costs of a non-compliant transaction are very low, the costs of complying with regulatory requirements might add a significant proportion to transaction costs, increasing the temptation to be non-compliant.

3.1.4. Diversity of preferences over time

The rental model means less commitment to a given asset. In choosing a car, a consumer might have a number of different potential situations in mind: moving furniture, where a van might be ideal; travelling short distances to access services, where a bike or an electric car might be best; and longer distances to visit family where a diesel car might be preferable.

If they are buying a car, they will have to strike a balance between those different potential uses, with limited information about their future frequency, whereas in a rental model they could hire a different car for each situation: a van from ZipCar to move furniture; an electric car from Autolib' or a rented bike for short trips in urban centres; a shared ride with BlaBlaCar to visit family; and hiring a car and driver with Uber to get home after a night out.

This diversity is likely to grow in importance with the development of the market, as smaller niches will represent more viable opportunities for sharing economy providers, and may also become more important as firms offer services through those platforms rather than peers (who are more likely to need a broad set of uses to justify ownership of an asset in the first place). It represents an alternative to mass customisation, satisfying greater diversity in tastes over time as opposed to greater diversity in tastes between consumers.

The potential welfare gains with greater variety in consumption can be substantial. The import of new varieties of goods from 1972 to 2001 is estimated to have increased US welfare by 3 per cent (Broda & Weinstein, 2004). Avoiding the need to compromise between different uses might

PE 558.777 I - 66

be somewhat more fundamental, reflecting an ability to gain additional functionality as well as a love-of-variety in itself.

3.1.5. Division of labour

Almost all examples of the sharing economy involve some element of a division of labour. It is most obvious in cases where labour is the primary service on offer. Platforms like TaskRabbit make it easier for tasks like the assembly of flat pack furniture, which might otherwise be undertaken by the consumer themselves or friends and family, to be done by a smaller group who find the task easier and can improve with greater practice. This is likely to lead to the tasks being completed more quickly and to a higher standard.

Other platforms where labour is a smaller component in the overall value of the good on offer, like car sharing, may also involve an element of division of labour. Owners of cars need a range of supporting services, either undertaking the work themselves or procuring it from others, such as insurance, maintenance and cleaning. Undertaking the work involves some skill and so does procuring such services from third parties, e.g. knowing where to get a good deal or what level of service is actually needed. The resulting division of labour could not only affect the quality-adjusted cost of those supporting services, it might also increase people's ability to focus on their own work and hence their productivity.

3.2. Providers

In our definition of the sharing economy, providers can either be peers who are sharing assets they would otherwise own and only use themselves or businesses (either corporate firms or individual traders) who offer their services. Providers can be expected to engage with the sharing economy to the extent that they are able to increase their earnings to an extent sufficient to offset any associated inconveniences, such as the transaction costs of offering their asset for sale, and risks, such as the potential the asset will be damaged.

Another distinction among providers, besides corporate form, can be borrowed from the entrepreneurship literature, between necessity and opportunity entrepreneurs (Block & Wagner, 2010). The impulse for a necessity provider is that they are responding to a problem largely external to the sharing economy. They may provide services on sharing economy platforms because they have become unemployed or their pay has been cut at work. The impulse for an opportunity provider is that they see an opportunity in the sharing economy, perhaps to supplement their existing income by making greater use of their existing assets.

The difference might be important with opportunity providers being more likely to drop out in the event that the conditions on offer with a sharing economy platform worsen, whereas necessity providers might instead be more likely to drop out if wider economic conditions change (e.g. an increase in vacancies in the wider labour market). Corporate providers, who if the offering from a platform worsens might have other opportunities including establishing new platforms, might generally be thought of as opportunity providers (though that might not be true of small firms).

PE 558.777 I - 67

3.2.1. Earnings and employment

Many providers will enjoy unambiguously higher aggregate earnings in the sharing economy than they would have otherwise, as they receive additional earnings alongside their prior income. Those renting properties when they were away for extended trips, which would otherwise have been empty, for example. Sperling (2015) suggests that renting properties on Airbnb might mitigate pressures on middle class incomes for those renting properties in the US, with typical hosts being paid $7,530 for renting accommodation for an average of 66 days a year.

There might also be opportunities for more people in some of these markets and in related markets as a result of the development of sharing economy platforms (Dervojeda, et al., 2013, p. 8):

'Apart from the direct creation of new jobs through the hiring of new staff, companies operating in the sharing economy also generate substantial amounts of indirect employment.'

On the other hand:

'It is self-evident that the benefits described above allow companies in the sharing economy to compete with conventional product service providers. This implies that some markets and jobs will therefore be destroyed.'

In some markets a reduction in the barriers to entry should mean the overall impact is an increase in employment, associated with a reduction in deadweight loss. For example, there are substantial differences in the number of taxis per resident in different cities (Bonamy, Dorso, & Guerrab, 2014). Increasing supply will mean more employment opportunities, as well as increased availability for consumers. If the number of taxis and chauffeured cars per 1,000 residents in Rome were to rise to the level in London, for example, then it would imply an additional 20,000 opportunities. Research for Uber has estimated that sharing economy transport services might create around 1,000 new full-time jobs in Stockholm (Stefansdotter, Utfall Danielsson, Nielsen, & Sunesen, 2015).

Table 3: Taxis and chauffeured cars, by city

Taxis and chauffeured City cars, per 1,000 residents New York 13.5 London 10.8 Stockholm 7.8 San Francisco 6.7 Madrid 5 Amsterdam 4.2 Paris (greater area) 3.4 Rome 3.2 Brussels 2.9 Berlin 2.6 Seattle 1.2 Houston 1

Source: Bonamy, Dorso, & Guerrab, 2014

PE 558.777 I - 68

Hall and Krueger (2015) find that variations in the volume of supply per driver do not affect average wage rates significantly, which implies that there will be a less of a threshold beneath which a driver is not working enough to justify their entry into the market (as their earnings per hour will equal those of drivers working longer hours).

The impact of growth in the sharing economy (and reductions in quality-adjusted consumer costs) on average earnings in a given sector is likely to depend on the source of the reduction in consumer costs. To the extent prices are lower for consumers because of higher utilisation or lower costs, then providers could still earn higher incomes. To the extent prices are lower for consumers because of increases in supply in industries where there were previously barriers to entry, the sharing economy might undermine existing occupations that are higher-paid.

If there is a large increase in the supply of taxis then, even if it leads to a substantial increase in aggregate earnings, there might be no individual taxi driver still able to command equilibrium earnings in the earlier constrained market.27 There have been suggestions that the final result of a removal of barriers to entry in markets such as road transport might be a very low level of equilibrium earnings, where workers are unable to 'earn even minimum wage' (Baker, 2014).

This kind of effect will have limits outside labour markets with very high unemployment (where increasing the volume of opportunities will have its own attractions). There are barriers to entry as a sharing economy provider, including the need for certain assets (e.g. a car, or human capital such as technical skills); the need to be able to maintain high ratings for consumer satisfaction (requiring scarce skills such as punctuality and politeness); and a geographical location near potential customers (a 'race to the bottom' normally requires competition with workers expecting much lower incomes in other economies).

The impact of the sharing economy is therefore likely to be an increase in aggregate labour earnings and an increase in individual earnings for new entrants to a sector, without which they would not enter. Hall and Krueger (2015, p. 23) find that the 'net hourly earnings of Uber's driver-partners typically exceed the average hourly wage of employed taxi drivers and chauffeurs', for example. While that comparison is not comprehensive, it does not include some costs and some taxi drivers and chauffeurs may enjoy other benefits, it suggests earnings in the sharing economy can match or better those in similar jobs, in similar industries. There may be a reduction in earnings for incumbent providers in sectors where barriers to entries led to earnings substantially above the new equilibrium price.

Opportunities for employment in the sharing economy may also increase over time. Employment in the sharing economy is likely to be lower to the extent that P2P models predominate. Households will provide most of the labour required themselves, using the shared asset. They are borrowing from a peer who is likely to have a similar level of skill and experience in using that asset. With B2C models, however, the work in preparing and maintaining the assets to be shared, or the services provided, is far more likely to be done as a job. People are employed to clean, maintain and administer shared cars and in some cases to drive the cars as well (giving the passenger the option of working during the journey).

27 This might not be the case if the barrier to entry removed itself reduced earnings, if it were some kind of cost which deterred some providers entirely and reduced the earnings of those who persevered.

PE 558.777 I - 69

In some cases, the sharing economy will see activity previously undertaken by the household itself (e.g. flat pack furniture assembly) carried out instead as a service for payment. In this respect, the sharing economy reflects a reversion of a pattern observed with Internet services where considerable value is created (e.g. with user-generated content on social networks) but not reflected in national income statistics (Brynjolfsson & McAffee, 2014, p. 115), earnings, employment or tax revenue (Cowen, 2011).

3.2.2. Other remuneration and benefits

There is considerable interest in the question of whether the sharing economy might be associated with a lower level of work-related benefits (Reich, 2015). Those benefits which might not be available to those working as self-employed providers in the sharing economy include: paid holiday, paid sick leave, employer pension contributions, maternity and paternity leave and employment protection. In the US, those benefits have been estimated at $20,423 a year for each worker (Brown, 2015), though a significant component in that total is healthcare coverage, which in the EU is generally not tied to employment. They may also not be subject to some restrictions such as working time regulations.

However it is important to bear in mind that the difference is a difference between self- employment and employment, rather than a distinctive feature of the sharing economy itself. The sharing economy is firstly not always an alternative to other employment. In those cases where the labour component is relatively small (e.g. the sharing of cars or accommodation) the impact on the general provision of such benefits is likely to be small. At the same time, in some of the services where sharing economy models are most developed, self-employment was already often the norm: taxi services, handyman services and domestic cleaning. Again the impact on the general provision of such benefits may be small.

In order for the sharing economy to have a significant effect on the provision of worker benefits, there would need to be either a large increase, driven by the sharing economy,28 in the prevalence of self-employment in a certain set of economic sectors, or an expansion of those sectors in which self-employment is the norm.

It is possible that self-employment might increase in its prevalence within sectors, as the reduction in transaction costs (particularly, the costs of finding customers) allows the self- employed to more easily participate in a wider range of markets. However the ability of firms to centralise marketing, billing or other such functions is only a part of their economic function and the extent to which staff are hired as employees may still vary based on other assets such as the employment of specific human capital. Generally, we can expect that the greater the asset specificity of human capital, the less likely self-employment will be viable, as both parties will prefer a corporate firm where they can establish the trust that their counterpart will not exploit the asset developed specifically for that transaction (Williamson, 1975). Firms are also a means of risk-pooling and workflow smoothing for employees relative to self-employment. The self- employed are more exposed to the 'feast or famine' problem of too much work or too little.

28 Self-employment has been rising even in advance of the development of sharing economy platforms, and there is even an extent to which the development of the sharing economy is a product/reflection of rather than driver of this trend.

PE 558.777 I - 70

It is possible that the increased prevalence of self-employment seen in the sharing economy is a temporary result of the peer-to-peer focus of many early platforms. To the extent that peer-to- peer transactions are displaced by business-to-consumer transactions over time, reflecting a division of labour and a reduced incentive to own assets that can more easily be rented, that is then equally likely to result in some reversion to the typical employment practices in the existing industries.

To the extent that benefits are diminished by increasing self-employment associated with the sharing economy, it is likely to create pressure for policymakers to find ways to extend those benefits to the self-employed (Sundararajan, 2015). This might involve recourse to public rather than employer provision; it might involve an extension of insurance and other financial markets; and it might be a consideration in the regulation of platforms. These questions will be considered in Chapters 6 and 7.

3.3. Competitors

There are broadly three types of market participant which at some level compete with the sharing economy and might therefore be affected by its development:

− Those providing substitute goods and services. − Those who have other uses for scarce assets. − Those manufacturing goods for ownership.

3.3.1. Other providers of goods and services

Existing services could clearly be affected by the reduction of barriers to entry and therefore more competition in the markets in which they operate. This has motivated some of the most forceful objections to the growth of sharing economy platforms, up to and including violent protests (Rubin & Scott, 2015). The impact can clearly be significant with a reduction in hotel revenue of 8 to 10 per cent in Austin, Texas, the market in which Airbnb capacity is greatest (Zervas, Proserpio, & Byers, 2015), and a particularly pronounced effect on lower-cost hotels not catering to business travellers.

From the perspective of the consumer, a reduction in prices and increase in the range of choices represents an improvement in their welfare. From the perspective of existing providers, who may have made substantial investments in developing capacity which is suddenly much less valuable, it is a significant adverse shock to their business. This might mean that transitional assistance is needed for those adversely affected. But general economic principles would counsel against attempting to frustrate a transition brought about by technological change.

An objection often cited by existing providers is that they are subject to a greater regulatory burden than those providing services in the sharing economy, that there is not a level playing field. Those offering accommodation in the private rental market might not have to provide the same levels of access for the disabled, for example, and drivers might not be subject to the same licensing requirements as existing taxi services.

PE 558.777 I - 71

To some extent, this lower regulatory burden might reflect a failure to enforce existing rules, possibly because chasing a large number of providers operating small businesses on sharing economy platforms seems disproportionate or infeasible to the relevant authorities. It may also reflect that those existing rules would be prohibitive for those providers, leaving them no realistic option but to shut down, and the authorities do not want to lose the potential benefits to consumers (particularly if the alternative might be illicit provision subject to less control than a provider using a sharing economy platform – see Figure 3, Figure 4 and Figure 5).

It is important to note, however, that a level playing field should not necessarily be a regulator's objective. It might be that less regulation is needed in some cases than in others and that represents a genuine efficiency gain from the technology (which regulatory policy should therefore not try to offset or frustrate). To give an extreme example, the regulatory burdens for nuclear power plants and wind turbines are not equal. There is normally a much greater level of safety regulation for the nuclear plant. It is possible to assert that that the nuclear power plant is therefore at an 'unfair' disadvantage, it is not able to compete on a level playing field with the wind turbine, but that would only really be the case if the burden of safety regulation were excessive relative to the safety risks posed by the nuclear power plant. It is also possible that the wind turbine is still facing an unfair disadvantage, if its lower level of safety regulation is, despite being lower, still disproportionate (and the nuclear power plant's regulatory burden is proportionate).

There are a number of reasons why regulation might not need to be the same for the sharing economy as for traditional providers.

The transparency provided by digital technologies might reduce the need for the regulation of sharing economy providers. If both the provider and the customer are registered by the platform, along with their bank account details, then it is harder for either party to abscond without fulfilling their responsibilities to provide the service and pay for it promptly. If providers are registered with the platform, then it might not be necessary for them to be registered with a local regulator. They are already not able to operate anonymously and could be tracked down in the event of something going wrong.

At the same time, assets provided on peer-to-peer platforms might already be covered by existing regulatory frameworks more suitable to small-scale provision. Disability access is already often a requirement of planning regulations for new domestic accommodation, for example, and that might mean there is already sufficient accessible property in a large and diverse range of properties available for short-term rental through sharing economy platforms. Additional requirements needed in a market with a relatively small number of large hotels might not be necessary.

It might be necessary to regulate prices for conventional taxis, as someone hailing a taxi on the street needs the reassurance of a meter that they will not be overcharged once they are in the taxi. It might not be necessary to regulate the prices of hiring a car and driver in the sharing economy though, as customers can see the price they will pay in the application29 before deciding to make the hire. The driver has no ability to artificially increase the charge once a passenger is in the car as charging is handled by the platform.

29 Including the degree to which any surge pricing might be in effect.

PE 558.777 I - 72

The objective should be proportionate regulatory interventions that address potential harms, which would otherwise not be addressed through self-regulation, rather than a superficial equality with the regulatory burden imposed on other economic activity.

3.3.2. Other uses for scarce assets

To the extent that assets are scarce and supply cannot increase to match demand, it is possible that platforms might intensify competition for them.

There are concerns in some property markets that the potential to let properties on Airbnb has led to a reduction in the supply of homes rented to those living in the city (Samaan, 2015).

It is unclear to what extent this problem is material. Research commissioned by Airbnb (GEWOS, 2014) found that in Berlin, a city in which this objection has been raised, around 0.5 per cent of the housing stock is rented on Airbnb in a given day and 0.06 per cent of the housing stock is rented on the platform for more than 120 days a year. It contrasts the around ten per cent of the around 9,500 properties rented on a given day to the rise in the Berlin population of 44,000 in 2013 and argues that Airbnb is a defensive strategy for those affected by rising rents (allowing them to offset some of that cost by renting their homes when they are away) rather than a source of the problem.

The share of the housing stock accounted for by property rented on Airbnb at any given time could rise over time. However the impact of that might not be to increase rents for residents (even leaving aside the potential for those residents to rent out their own properties). If the source of the scarcity is that the amount of development overall is limited (by space or by the scale or volume of construction that residents will tolerate) then, by limiting the pressure to build more dedicated hotels, sharing economy platforms could allow for more private accommodation to be built. If sharing economy platforms result in a rise in utilisation of the housing stock as a whole, then they are likely to exert a downward pressure on rents for private accommodation, not an upward pressure.

There might be some cases where very strong tourist demand and very strong limits on potential development (e.g. Venice) mean that this logic does not hold and protection from competition with tourists lowers rents for residents. In some such cases, it might be that tourist taxes (which are already collected by Airbnb in some jurisdictions) or other regulatory interventions become necessary.

Another case where there are concerns that the sharing economy might increase competition for a scarce asset is roads. There have been concerns in London that an increase in the prevalence of minicab services, with the advent of services like Uber, creates congestion and lowers journey speeds. The Mayor of London has proposed a cap on the number of minicabs. Uber contests whether this is the case, arguing that average speeds have risen for its drivers (Cellan-Jones, 2015). Uber has also commissioned research from Copenhagen economics which finds that sharing economy transport services will lower 'traffic intensity and congestion' (Stefansdotter, Utfall Danielsson, Nielsen, & Sunesen, 2015), due to a reduction in the number of daily car trips.

PE 558.777 I - 73

Over time, the result is likely to depend on the alternatives displaced by sharing economy platforms:

− To the extent that sharing economy platforms substitute for private car or taxi use, they are likely to reduce congestion. They reduce the need for cars to spend often significant amounts of time searching for parking. The average time spent searching for parking reported in an IBM survey of 20 major global cities was nearly 20 minutes (IBM, 2011). Over time platforms can also encourage pooling services so that more people are travelling in a given car, further increasing utilisation and therefore provider revenues. They are also likely to spend less time waiting for fares than conventional taxis, as utilisation tends to be higher. − To the extent that sharing economy platforms substitute for walking, cycling or public transport, or increase the number of journeys overall, they will increase congestion. This may occur to some extent, however car sharing has been found to be associated with increased walking and use of public transport (Shaheen, Cohen, & Chung, 2009). Car and driver services like Uber can enable the use of public transport for some consumers, by accounting for the start or end of journeys where the bulk of the distance is accounted for by public transport; and barriers to cycling can be reduced by cycle sharing, another part of the sharing economy.

It seems unlikely that the sharing economy will increase congestion over the medium- to longer-term, as there is a strong incentive to maximise utilisation.

Finally, in some cases, it is possible that consumers might lose out in the event that it is possible for sharing economy providers to trade a claim on an asset provided at a discounted price. Instead of being able to buy it more cheaply directly themselves, customers might be forced to trade through a middle-man. While this constitutes a relatively special case, there are examples of it occurring, for example with services in the United States where people pay for someone else to vacate an occupied parking space (Munger, 2014):

'[There] are people now who instead of sitting at home playing video games, drive their old junker jalopy across the bridge from Oakland and look around for parking spaces. And if they see one, they jump into it; and they pay the minimum amount and then put themselves down. MonkeyParking App has two buttons. One is: I have a parking space; and it looks for your location, because it's a cellphone app. And the other is: I want a parking space; and it looks for your location. And then it matches people that have them with people that want them. And so-- Russ: It encourages staking a claim. It's like people who go out and get a good URL (Universal Resource Locator), hoping that someone will also want it; and then they can re-sell it. So it encourages people to prospect for parking spaces--and make it even harder to find one. [Munger]: It makes it impossible. It's literally impossible to find a parking space.'

It is debatable whether this example should count as a part of the sharing economy, as people are paying for access to an asset which the provider does not themselves own. Those using the platform are exploiting the pricing of parking below market rates. Consumers are willing to pay more for it than the actual amounted charged by local authorities. It is therefore worthwhile to expend resources to stake a first-come, first-served claim. Local authorities could raise the price to a market rate at which it is no longer worthwhile, if they wished to capture the benefits being redistributed by the application.

PE 558.777 I - 74

3.4. Other policy priorities

3.4.1. Manufacturing

Manufacturers have the potential to be affected significantly by the growth of the sharing economy. The most direct effect might be that if assets are used more efficiently, there might be less demand and therefore volumes might fall significantly. That might be a challenge for established manufacturers, leading to excess capacity.

Manufacturers might gain, however, if they are able to either deliver a higher-value product or offer associated services.

If volumes fall significantly, but assets are being used a lot more, demand for reliability, durability and quality might rise significantly. The increased utilisation associated with sharing would dilute the fixed cost of purchase over more use and increases the need for durability. This is already the case in bikes, where the bikes used 'are proprietary designs and tend to be heavy due to their robust, comfort- and style-oriented design' (ITDP, 2013, p. 76) This might mean that the share of the premium market rises in cars, for example, a market in which European manufacturers are particularly competitive. European exports of cars to the United States, China and Japan – the largest international markets – are all dominated by the premium segment (DG ENTR, 2014).

Manufacturers might further respond to the changes brought on by the sharing economy by creating platforms of their own in order to share their products directly with consumers. This is already occurring to some extent with BMW launching the car sharing platform DriveNow, for example, and Ford trialling Ford2Go (and partnering with ZipCar in the United States).

3.4.2. Income and wealth inequality

There are concerns that the sharing economy might exacerbate inequality. Existing owners of assets will be able to increase the return on those assets by increasing their utilisation (homeowners can rent out their homes using Airbnb, for example), but fewer people will own assets.

The impact of the sharing economy seems more likely to be a reduction in inequality.

First, the benefits of sharing economy platforms to consumers, such as reduced costs and increased access (leading to earning power if they can reach a wider range of potential employers) could increase the welfare of low-income consumers and their ability to save. Studies using data from existing P2P platforms have found that 'below-median income consumers will enjoy a disproportionate fraction of eventual welfare gains from this kind of 'sharing economy' through broader inclusion, higher quality rental-based consumption, and new ownership facilitated by rental supply revenues' (Fraiberger & Sundararajan, 2015).

At the same time, if low-income households no longer need to borrow to finance consumer durables, or use their savings to purchase such durables, they can instead invest in financial assets which generate higher returns.

PE 558.777 I - 75

If under-utilised assets are used more and fewer assets therefore need to be bought, that could also reduce the demand for capital, reducing returns on capital (affecting those with the greatest existing stocks of wealth). While the returns on some assets people own already might rise as they increase their utilisation, the impact of the sharing economy might therefore be to reduce returns across the capital stock as a whole (it has become less scarce and the cost of renting it is therefore lower).

The inventors and owners of the platforms themselves may, in some cases, enjoy large returns, but past experience suggests that those returns may only be a small share of the overall welfare gains. Earlier research has found that 'only a minuscule fraction of the social returns from technological advances over the 1948-2001 period was captured by producers, indicating that most of the benefits of technological change are passed on to consumers rather than captured by producers' (Nordhaus, 2004).

To the extent that the welfare gains from the sharing economy accrue most to those with limited access to valuable assets now, it is likely to reduce inequality. Beyond that, it might reduce the salience of wealth inequality, by reducing the degree to which either wealth or the ability to borrow is necessary to access valuable assets. If it is possible to access assets when you need them, without saving or borrowing money to buy them outright, then it matters less whether you have that money saved in the first place.

3.4.3. Social exclusion

Whilst there has been considerable attention paid to the question of whether ratings systems and other measures are sufficient to establish consumer trust, or keep consumers safe, much less attention has been given to the possibility of ratings systems creating a new form of social exclusion. Any reputational system is likely to include the possibility that the ratings of certain consumers or providers become so poor that they cease to be able (or even in some cases eligible) to participate in the market at all.

A consequence of this could be that they become unable to rebuild their reputations and so, once excluded, are permanently excluded. If the sharing economy becomes a large share of the economy as a whole, exclusion from sharing economy activities could mean social exclusion from large part of economic and social life. This might particularly be the case in economies where unemployment is high and there are fewer other opportunities for potential sharing economy providers, as platforms will face less pressure to accommodate providers in order to be able to match consumer demand.

There might also be exclusion on the demand side. Mistakes which are then difficult to erase with subsequent good behaviour and which mean someone is unable to access services. This is a limited problem so long as there are either other options outside the sharing economy, where there is no rating system, or there is competition among sharing economy platforms using different rating systems. If the sharing economy were to dominate certain activities and there were to be sharing of information across ratings systems then, while that might increase safety, it might also increase the risk of participants being frozen out of markets.

PE 558.777 I - 76

These problems might be enhanced if there is scrutiny of a potential market participant before they have even engaged with a platform. This could be based on formal prior checks (e.g. using credit checking services) or could be based on consumer choices using the information provided through the platform. Discrimination might take place without any intention on the part of the platform which exacerbates social exclusion.

One study found that what it referred to as 'non-black' providers using Airbnb charge approximately 12% more than what the study referred to as 'black' hosts for an equivalent rental (controlling for all information visible in the Airbnb platform) (Edelman & Luca, 2014). These lower prices may reflect actual or perceived-by-the-provider discrimination by consumers. Other research argues out that price-setting may reflect a complex range of motivations, with a lower price likely allowing a provider to choose between a wider range of potential consumers, for example (Ikkala & Lampinen, 2014). It seems equally possible that the price charged reflects the expectations of a reasonable price, which might vary between different communities for a range of reasons. Despite those reservations, however, it still seems plausible that there might be discrimination in the choices people make on the basis of the information provided in sharing economy platforms with the aim of encouraging a feeling of safety and community.

In some markets, the sharing economy might reduce social exclusion by increasing access to goods and services. Increased access to mobility as a service, for example, might reduce isolation among the elderly. There is evidence that Uber has extended transport services to low income areas served poorly by established taxi services (Smart, et al., 2015). If the ability to access these markets becomes increasingly essential, however, and platforms and market participants are extremely risk averse, then it might be difficult for those who do not appear reliable to those participants to establish themselves in the market.

3.4.4. Minimum wage laws sharing economy platforms, like other sources of work, may not provide an income felt to be adequate for all participants. In those markets where the skills required are modest (e.g. domestic cleaning) and there are a limited number of other vacancies, workers might accept low pay. This does not mean that the platforms are bad for those workers in themselves, as the alternative may be other even lower income work or unemployment, but it might mean they undermine broader attempts to raise wages.

A report for the UK Government called for platforms to offer a Living Wage, a wage sufficient for people to achieve a reasonable standard of living without recourse to benefits (Wosskow, 2014), and recommended that be required as part of a quality kitemark for platforms. The limitation on these efforts is that they might simply redirect work to other sharing economy platforms or other markets entirely if consumers prefer lower prices. They may also be simply inappropriate for some workers — the rationale for requiring a Living Wage for those that are topping up a main income is unclear.

Another response to low earnings for some workers, on some platforms, as lower transaction costs and more efficient markets cause earnings to trend to the marginal product of labour, might be encouraging training that allows workers to transition to higher-skilled and better-

PE 558.777 I - 77

remunerated work. Lower barriers to entry generally encourage investments in training (Bassanini & Brunello, 2010) and over time platforms might increasingly offer training as a way to attract and retain workers and as platforms increase their ability to offer more valuable services to consumers. This might be particularly likely in labour markets where there is a higher vacancy rate, where the pressure to retain workers is greater. The importance of labour market institutions conducive to high employment might therefore rise.

3.4.5. Combating tax evasion and other data-intensive tasks

There have been concerns that the sharing economy might be associated with a reduction in government tax revenues (Baker, 2014). To the extent this is a concern about providers not being legally required to pay certain taxes for which established providers are liable (e.g. business property taxes) then it might imply that the growth of sharing economy platforms will increase pressure for reforms to Member State tax policies.

To the extent that it is instead a concern about tax evasion, e.g. providers not reporting their income, then growth of the sharing economy might become an opportunity to significantly improve compliance if platforms were to report the data they hold on transactions (as already happens in Amsterdam and some US jurisdictions (Sullivan, 2015)). Tax compliance is generally stronger in economies with a higher penetration of electronic payments. Cash transactions create less of a record and are therefore more likely to be under-reported. Schneider (2013) reports that:

'Paying with cash makes it easier to engage in the shadow economy, since cash payments cannot be traced. The shadow economy is clearly a cash-based economy, and cash is the fuel in its engine.'

The same report finds that the Shadow economy accounted for 18.4 per cent of GDP in the EU27 in 2013, with a range from 7.5 per cent in Austria to 31.2 per cent in Bulgaria. The wholesale and retail trade and hotel and restaurant sectors have a larger than average shadow economy share at around 20 per cent of GDP. In the transport, storage and communication sector, the shadow economy accounted for 15 per cent of GDP (and this will include public transport where the shadow economy is likely to be much less prevalent). Those are sectors where firms interact directly with consumers and where individual transactions are often small and sharing economy platforms can compete, displacing cash transactions with recorded electronic transactions which could therefore be automatically reported to revenue authorities.

A theoretical limitation on this facility for governments could be if the tax or regulatory burden reached a point where it became worthwhile for providers and consumers to use a sharing economy platform to find each other, cancel the transaction and then go ahead on a cash basis (avoiding the transaction being reported for tax purposes). The platforms have a strong incentive to avoid this as it would also remove their return on the transaction.

There might also be a range of other uses that governments could make of the data generated by sharing economy platforms in fields such as urban planning or even disaster response. sharing economy platforms might be resistant to sharing any data though, out of concern that the data will be used as a means to impose limits on their growth. They might have a concern that the data would be used to attempt to target providers and force them out of their market.

PE 558.777 I - 78

While governments could in theory compel the release of valuable data, platforms might then decide that the market is not one in which they wish to operate (data protection rules might also be breached by any release of data not accepted in advance by the provider and/or consumer).

A stable regulatory framework for the sharing economy might therefore have the added benefit that it could effect a valuable transfer of data to public bodies can take place. The challenge could essentially be to establish a commitment mechanism where governments commit to allow the growth of sharing economy platforms and those platforms share the data which will maximise the potential benefits to governments.

4. Quantifying the impacts of the sharing economy

The scale of the impacts of the sharing economy, including its positive and negative externalities, will, in part, depend upon its growth. Its growth will, in turn, depend upon the obstacles that will be considered in the next chapter. At this stage, however, our objective will first be to construct an approximate estimate of the extent to which the principal relevant assets are currently under-utilised and the value of increasing that utilisation to 100 per cent.

There are, however, a wide range of potentially-relevant assets and it is neither practical nor necessary to estimate utilisation for all of them. The most valuable non-financial assets held by most households will be their labour, their house, their car. We will therefore estimate the under-utilisation of labour, accommodation and cars directly and assume the utilisation of other assets to zero. We therefore implicitly assume that the main impact of the increased utilisation of other assets will be as a means to increase the utilisation of labour (for example, while a service that allowed the sharing of domestic textiles would require labour to produce, distribute and clean those domestic textiles).

It is important to note, at this stage, that we are not suggesting that the sharing economy will lead to the complete utilisation of these assets. There are obstacles to its growth which are difficult to overcome (not all of which relate to policy choices). This is particularly true in the labour market where we essentially quantify the value of ending unemployment and under- employment. However, understanding the scale of the under-utilisation of assets which might be provided through sharing economy platforms, including human capital, provides a first step to understanding the value of removing any obstacles and illustrates at a high level the scale of the opportunity.

4.1.1. Under-utilisation of labour

The under-utilisation of labour can be thought of simply as unemployment and under- employment,30 multiplied by the value of that labour. There is no direct evidence for the value of that labour as the employees concerned are not employed and are therefore not receiving a market wage. In order to produce a conservative estimate, we will therefore assume that the additional earnings were that labour to be employed reflects the gross earnings of a single

30 There are other supplementary indicators to unemployment which might also indicate available labour, particularly persons seeking work but not immediately available and persons available to work but not seeking work. For the sake of a conservative estimate, we will exclude those categories (as it is difficult to ascertain the potential for inactive workers to re-enter the labour market with enhanced opportunities).

PE 558.777 I - 79

person earning 50 per cent of the average.31 In this simple estimate, we assume that the labour of someone unemployed is half that if they were employed, as there is some utility in time spent not working, and that labour of someone under-employed is one quarter employed (they are half-employed and the remaining time has fifty per cent of the value of employment).

Table 4: Under-utilisation of labour

Active Under- Un- Under- Assumed population, utilisation, employment employment earnings, € 000s 2013, €m ((A/2)+(B/4)) * A B C D C * D EU28 11% 4% 241,680 19,597 309,042 Belgium 8% 3% 4,947 29,125 7,240 Bulgaria 13% 1% 3,371 2,937 668 Czech Republic 7% 1% 5,307 7,705 1,503 Denmark 7% 3% 2,890 26,530 3,201 Germany 5% 4% 41,713 26,658 40,588 Estonia 9% 1% 680 7,860 246 Ireland 13% 7% 2,163 17,986 3,210 Greece 28% 4% 4,843 13,234 9,519 Spain 26% 7% 23,190 16,904 57,724 France 10% 5% 28,766 21,870 40,578 Croatia 17% 2% 1,842 7,052 1,182 Italy 12% 3% 25,259 19,838 33,447 Latvia 12% 3% 1,014 5,781 396 Lithuania 12% 2% 1,465 5,052 481 Luxembourg 6% 2% 254 30,034 259 Hungary 10% 2% 4,334 6,198 1,511 Malta 6% 3% 188 11,292 82 Netherlands 7% 7% 8,932 26,316 12,458 Austria 5% 4% 4,336 27,078 4,286 Poland 10% 2% 17,361 5,765 5,680 Portugal 16% 5% 5,285 10,915 5,451 Romania 7% 3% 9,202 3,740 1,437 Slovenia 10% 2% 1,008 10,259 581 Slovakia 14% 2% 2,716 6,561 1,336 Finland 8% 3% 2,676 25,946 3,368 Sweden 8% 6% 5,116 29,772 7,958 United Kingdom 8% 6% 32,393 22,567 38,743

Sources: A: Eurostat, Unemployment rate by sex and age groups - annual average, % [une_rt_a] B: Eurostat, Supplementary indicators to unemployment by sex and age groups - annual average, 1 000 persons and % [lfsi_sup_age_a] C: Eurostat, Population, activity and inactivity - annual averages [lfsi_act_a] D: Eurostat, Annual net earnings [earn_nt_net]

31 Cyprus has been excluded from this analysis, as earnings information was not available, but it shouold be included in the EU28 aggregate values reported.

PE 558.777 I - 80

The value of the under-utilisation of labour across the EU28 is therefore €309bn on this estimate. The highest value is in Spain at €58bn, reflecting a high unemployment and under-employment rate in a large economy.

4.1.2. Under-utilisation of accommodation

There are a number of reasons why accommodation might be under-used, not all of which are likely to be remedied by the sharing economy, for example:

− People leave the home during the day, but usage will peak overnight. − People might prefer a larger property than they need most of the time, so that they can accommodate visitors (e.g. parents whose children have left home). − Disparities in regional economic performance and the long life-span of much residential accommodation might mean that there are properties empty in areas where there is low demand because there are few economic opportunities in the surrounding area.

We therefore consider the extent of the under-utilisation of accommodation on the basis of the frequency with which people take trips. This means we focus on under-utilisation of property related to its occupiers leaving a property that would otherwise be occupied (for which we can therefore surmise there is some degree of demand). In order to filter out short-trips, where it might often be impractical to find an alternative use for a property, we restrict to trips of 4 days or more. We then compare that to a theoretical maximum of the population multiplied by 365 to establish under-utilisation, the share of days in which people are away from home.

Table 5: Under-utilisation of accommodation

Trip days, Population, Population days Under-utilisation 2013 2013 A B C = B * 365 D = A / C EU28 4,926m 507m 185,114m 3% Belgium 86m 11m 4,074m 2% Bulgaria 15m 7m 2,659m 1% Czech Republic 93m 11m 3,838m 2% Denmark 70m 6m 2,045m 3% Germany 1,049m 82m 29,938m 4% Estonia 8m 1m 482m 2% Ireland 49m 5m 1,676m 3% Greece 55m 11m 4,038m 1% Spain 473m 47m 17,056m 3% France 1,035m 66m 23,936m 4% Croatia 42m 4m 1,556m 3% Italy 296m 60m 21,785m 1% Cyprus 15m 1m 316m 5% Latvia 10m 2m 739m 1% Lithuania 16m 3m 1,085m 1% Luxembourg 9m 0.5m 196m 5%

PE 558.777 I - 81

Trip days, Population, Population days Under-utilisation 2013 2013 Hungary 40m 10m 3,617m 1% Malta 3m 0.4m 154m 2% Netherlands 225m 17m 6,125m 4% Austria 81m 8m 3,085m 3% Portugal 57m 10m 3,828m 2% Romania 52m 20m 7,307m 1% Slovenia 13m 2m 751m 2% Slovakia 25m 5m 1,975m 1% Finland 79m 5m 1,981m 4% United Kingdom 718m 64m 23,322m 3% Switzerland 107m 8m 2,934m 4%

Sources: A: Eurostat, Number of nights spent [tour_dem_tntot] B: Eurostat, Main scenario - Population on 1st January by sex and single year age [proj_13npms]

The average under-utilisation for the EU28 is 3 per cent.

The main approximations involved in this method are that it focuses entirely on capacity that is unused because the occupier is away, rather than capacity under-used because the normal occupation of the property is incomplete. Holiday homes, for example, are not included which may be empty most of the year and rented through sharing economy platforms. Equally people may share spare rooms on sharing economy platforms (or space in their living room). On the other hand, the model assumes that there is a space available if someone is on a trip, where it is possible that, for example, one half of a couple might take a trip but the other half still needs a room. It seems a reasonable conservative approximation, however. The final result should be a reasonable conservative estimate.

While a home might need more cleaning or maintenance if it is occupied more, it will not depreciate faster with 3 per cent higher occupancy. We can therefore work on the assumption that taking up this under-utilisation would represent a saving.

Consumption in sectors where the main relevant asset is the principal residence represents total final consumption of €2,600 per person,32 or €1.3trn in total, across the EU28. 3 per cent of that value lost to under-utilisation is around €35bn in annual consumption.

4.1.3. Under-utilisation of cars

In order to estimate the under-utilisation of cars, we multiply the peak percentage of the population travelling (normally the number travelling in the morning rush hour) by the modal share of road transport and the population for each Member State. We then compare that to the number of private cars in that Member State. Unfortunately, time use data is only available for a selection of Member States, plus Norway, but there are enough for a reasonable sample.

32 Actual rentals for housing: €700; Imputed rentals for housing: €1,800; Maintenance and repair of dwellings: €100.

PE 558.777 I - 82

Maximum Passenger Owned cars, Under- population Population Required Excess car modal 2011 or utilisatio travelling, , 2013 cars cars share, 2013 2012 n 2000 D = A * B F = D - A B C E F / E * C E United Kingdom 8% 86% 63.9m 4.4m 28.5m 24.0m 84% Slovenia 8% 86% 2.1m 0.1m 1.1m 0.9m 86% Poland 6% 80% 38.5m 2.0m 18.8m 16.8m 90% Norway 8% 90% 5.1m 0.4m 2.4m 2.1m 85% Latvia 14% 77% 2.0m 0.2m 0.6m 0.4m 63% Lithuania 12% 91% 3.0m 0.3m 1.7m 1.4m 81% Italy 10% 80% 59.7m 4.6m 37.1m 32.5m 88% France 11% 85% 65.6m 6.2m 32.6m 26.3m 81% Finland 7% 85% 5.4m 0.3m 3.0m 2.7m 89% Spain 9% 81% 46.7m 3.3m 22.2m 19.0m 85% Estonia 11% 84% 1.3m 0.1m 0.6m 0.5m 80% Germany 6% 86% 82.0m 4.3m 43.4m 39.2m 90% Bulgaria 12% 81% 7.3m 0.7m 2.8m 2.1m 75% Belgium 10% 77% 11.2m 0.8m 5.4m 4.6m 84%

A: Eurostat, Participation rate in the main activity (wide groups) by sex and time of the day - collection round 2000 [tus_00startime]. This breaks down the day into ten minute periods and asks about the main activity in which people are engaged at that time. B: Eurostat, Modal split of passenger transport [tran_hv_psmod]. The modal split given in the data is between passenger cars, trains and motor coaches, buses and trolley buses. We take passenger cars as a percentage of the total. This seems likely to lead to a conservative final estimate of excess cars, as many of those travelling at rush hour are likely to be walking. C: Eurostat, Main scenario - Population on 1st January by sex and single year age [proj_13npms] E: Eurostat, Passenger cars, by motor energy [road_eqs_carmot]. The latest data available was used. This related to 2012 for all of the Member States listed above, and Norway, except the United Kingdom, Lithuania and Belgium, where it related to 2011.

The results suggest that the number of cars could be reduced by between 63 per cent and 90 per cent, with the largest potential reductions in Germany and Poland. The average under- utilisation across this sample, weighted by number of cars owned, is 86 per cent. The most recent estimate for the number of cars owned across the EU28 is an ownership rate of 455 per 1000 inhabitants in 2006. If we apply that to the EU28 population in 2013 and the ownership rate above, that would imply a potential reduction in the number of cars needed of nearly 200 million.

This method is not perfect, in particular it does not account for the need for some cars to be travelling between different journeys, or out of use for maintenance.33 There will be some necessary spare capacity. At the same time, and potentially more importantly there is clearly the

33 Bear in mind, however, that this estimate is based on peak demand. It is therefore not assuming that cars are running 24 hours a day, there will still be considerable redundancy outside peak hours, and simply assumes that they are all carrying a passenger at the peak, which in most Member States is somewhat before 8.00am. This is also relevant to the question of how much more quickly those assets are likely to depreciate.

PE 558.777 I - 83

potential, particularly at peak times, for multiple passengers in a single car. We therefore considered an alternative estimate where peak demand represents half of the peak number travelling (two passengers per car on average at peak), but twenty per cent of the car stock is not carrying a passenger. In that scenario, the weighted average potential reduction in car numbers rises to 91 per cent.

These estimates should be considered indicative, but they indicate that the ultimate potential of the sharing economy, were the obstacles described in the next chapter removed, might be a reduction in car ownership on the order of 90 per cent.

It should be noted, however, that cars are likely to require more maintenance and depreciate more quickly (with more frequent replacement) if they are being used more intensively. If cars used in the sharing economy had to be replaced three times as often, for example, then the true reduction in car purchase costs might instead be around 60 to 70 per cent.

Consumption of cars is around €500 per person, or €254bn in total, across the EU28. To take the lower end of the estimated range at 60 per cent, we can therefore estimate a potential under- utilisation of €152bn in annual consumption.

4.1.4. Other sectors

If we remove those sectors where accommodation (and furniture), cars or human capital are not one of the principal assets, then the sectors left are household textiles; household appliances; tools and equipment for house and garden; other major durables for recreation and culture; and newspapers, books and stationary. The combined value of these sectors is €500 per capita across the EU28, the same as the purchase of vehicles.

It does not seem proportionate to estimate for each of these sectors separately, and any estimate is necessarily an approximation in any case as there are elements in other sectors which might be subject to sharing economy provision. If, as one scenario, we assume that these additional sectors are subject to under-utilisation of 30 per cent (half that for cars) it implies an aggregate under-utilisation across the EU28 of €76bn. If, as an alternative scenario, this under-utilisation assumption were reduced to fifteen per cent, that reduces the overall estimate of the potential increase in utilisation associated with the sharing economy by €38bn, i.e. by around 7 per cent.

The welfare loss from the under-utilisation of labour, accommodation, passenger cars and in other sectors is therefore €572bn, based on the calculations and assumptions described above. Of those components, accommodation is very valuable but better-utilised at present. Passenger cars are less valuable than accommodation but subject to a much greater degree of under- utilisation. Labour under-utilisation is the largest component, the most valuable opportunity if utilisation can be increased by sharing economy platforms. Later in this report, we will consider the extent to which that under-utilisation might be overcome in practice.

PE 558.777 I - 84

Chapter 4 - Are there obstacles or barriers which prevent the sharing economy from reaching its full potential?

KEY FINDINGS

- The need for a certain level of digital access and skill is currently a significant obstacle to the sharing economy but one expected to decline in importance rapidly. - Physical barriers to participation in the sharing economy are significant but may be overcome by new business models. - Consumer preferences for ownership are a significant obstacle to sharing, but can be reduced in importance as features of sharing economy markets now such as product scarcity risk and a lack of diversity in products on offer become less pronounced over time. - In some economies, labour market obstacles, e.g. skills mismatches, will inhibit the growth of the sharing economy. - The need to establish trust is a key challenge for the growth of the sharing economy, but one that platforms can meet over time in a range of different ways. - Tax and other policy choices not intended to affect the sharing economy might still affect its growth in each economy. - Regulation can deter sharing economy growth through outright bans, regulatory costs which deter self-employment, regulatory costs which deter marginal transactions or through inconsistencies and idiosyncrasies in intellectual property rules.

The scale of the sharing economy will ultimately depend on the extent to which there are barriers to its expansion in the sectors where platforms already operate, or its expansion into new sectors. These barriers include fundamental obstacles, such as the penetration of technologies like smartphones increases, which might change over time; institutional features that might affect the ability of the sharing economy to grow in some Member States such as rigid labour markets; and specific regulatory frameworks that might support or inhibit the growth of the sharing economy.

While policy might affect the salience of any of these barriers, we consider them in order of how fundamental they are and the extent to which they represent specific policy choices rather than features of a Member State or its well-established economic institutions:

− Digital access and skills. − Physical barriers. − Consumer preferences. − Labour market obstacles. − Trust.

PE 558.777 I - 85

− Tax and other policy. − Regulation.

The value of these obstacles and barriers will be quantified in the next chapter.

1. Digital access and skills

In order to engage with the sharing economy, market participants need to be comfortable with the technologies used. In most cases, this means a smartphone and the platform's application. Smartphone penetration is substantial in most Member States and rising (the fall in the minimum in 2012 represented the addition of several Member States, particularly Romania with the lowest smartphone penetration, to the sample).

Figure 6: Figure 4.1: Smartphone penetration, by Member State, 2013

Source: Google, Our Mobile Planet

If smartphone penetration continues to rise to 90 per cent or more, which is expected to happen in some Member States by 2018 (Arthur, 2014), then digital access and skills will become a less significant obstacle to the growth of the sharing economy. Pending that, however, access to smartphones is quite a fundamental and widespread obstacle to the sharing economy.

It might be possible that many users own those phones but principally use them for calls and older features such as email and text messaging, but it seems that most smart phone owners are regular users of applications (and are not prevented from doing so by external limitations such as a lack of network coverage). Given that sharing economy platforms have every reason to make their applications as easy to use as possible, and to market them to new potential customers, it seems unlikely that the ability to use the applications will be an obstacle in itself, for those consumers who have and use smartphones.

PE 558.777 I - 86

Figure 7: Frequency of use of smartphone applications (social networking, camera, gaming, etc.) excluding email/messaging, 2013

Source: Google, Our Mobile Planet

2. Physical barriers

There are a number of geographical and other physical barriers which might limit the development of the sharing economy, including:

− Low population density. − High costs of transport for sharing.

2.1. Low population density

Most sharing economy platforms are currently focused on serving customers in urban areas. This may in part be a coincidence, reflecting the priorities of founders who were themselves often based in urban areas. It may also reflect in part that there are advantages to sharing economy services that are less salient in suburban or rural areas. The reduced need for storage

PE 558.777 I - 87

space, for example, might be less of an advantage in suburban and rural areas where space is less expensive.

The most important reason why the sharing economy is likely to develop first in urban areas, however, is that higher population densities will naturally make it easier to achieve the critical mass a platform needs. If more people live within a given distance, then there will be a larger group between which assets can be shared. That will naturally improve the economics of the sharing economy.

This barrier might decline in importance over time if platforms can improve the matching of consumers and providers or if the penetration of sharing economy provision rises overall. Suburban communities and small towns, for example, might see growth in the provision of those sharing economy services that reach saturation in urban markets. Platforms may also grow which cater particularly to rural or suburban communities.

2.2. High costs of transport for sharing

There are some goods and services where transport costs make sharing economy services prohibitive. Many consumer durables, for example, are very heavy and could not readily be shared.

In some cases, that might be overcome by not sharing the asset itself (e.g. a heavy washing machine) but instead providing a service using that asset (cleaning the clothes) and thereby displacing the need to own an asset left idle most of the time. As the clothes are easily-moved, that makes it more practical to clean them using a shared asset. Digital platforms and portals which connected existing cleaners to customers and made service washing more practical might thereby extend the rental model, creating a sharing economy service.

In other cases, though, that will not be possible as the transport costs are high whether or not the asset itself is being moved. Dishwashers, for example, will be idle for most of the day in most homes, but unless their design were to change dramatically, they will not be transported themselves and the weight and fragility of the dishes would preclude their being transported either. We can probably therefore say that dishwashing is a service that is unlikely to become a part of the sharing economy. There will be other similar cases.

3. Consumer preferences

While sharing economy platforms might make rental models more practical as an alternative to ownership, many consumers might still prefer ownership, either in itself or because of remaining advantages to owning an asset.

3.1. Ownership

Customers may have a pure preference for ownership. This can be overstated and an observed preference for ownership might reflect other factors such as an expectation that the asset will

PE 558.777 I - 88

appreciate in value, but attachment to ownership can still clearly exist. It is likely to be particularly prominent with goods which have an aesthetic value, which can be prized as special. We might therefore expect the sharing economy to have less potential for those goods where there is an attainable aesthetic component. Manufacturers who wish to encourage the broad ownership of shareable goods might do so by making the goods more varied and aesthetically-attractive.

It is possible that a pure preference for ownership might, at least in part, be a temporary result of people being accustomed to owning certain assets. If it becomes more practical and therefore more common to hire those assets then, over time, a preference for ownership might decline. Renting might become the norm for assets people would previously have expected to own. Aesthetics might become more standard for most assets, with personalisation focused on a handful of owned devices (including the smart phone or watch or other device used to access sharing economy platforms).

Products might also be adapted to sharing, with manufacturers eager to capture sharing economy markets reducing the technical costs of sharing, a consumer needing to learn how to use a new asset each time they hire it (Lamberton & Rose, 2012), by simplifying and standardising how they are controlled, for example.

3.2. Diversity of preferences between consumers

People might customise assets they own for practical reasons (a car seat added to a vehicle) or for aesthetic reasons (large wheels). While it may be possible to customise shared assets to some extent, those customisations would either need to represent tastes shared by a sufficient number of other potential users (and an increased diversity of products would increase scarcity risk), or they would need to be made and then undone at the beginning and end of each use.

Customisation is widely understood to be increasingly important in manufacturing industries, with a trend towards Mass Customisation. There are understood to be four types of mass customising firm (Gilmore & Pine, 1997): collaborative customisers, who work with their consumers to design a good which fits their needs; adaptive customisers, who produce a standard good which consumers can then customise after market; cosmetic customisers, who produce a standard product but present it in different ways to different consumers (a standard t-shirt with a different logo printed on it); and transparent customisers, who 'inconspicuously customize their offerings within a standard package' without any explicit statement of preferences from their customers (who simply experience it as a service they prefer to other offerings).

It may be possible to offer minor adaptive, cosmetic and adaptive customisations through a sharing economy platform, but collaborative customisation or more substantial customisations seem unlikely. The attraction of customisation is therefore likely to remain a barrier to the sharing economy. Whereas differences in a given consumer's tastes over time can more easily be accommodated by the sharing economy (consumers can hire a different asset each time), differences in tastes between consumers is harder to accommodate than in a conventional ownership model.

PE 558.777 I - 89

3.3. Product scarcity risk

If someone purchases a rivalrous asset (like a car) it is theirs to use when they want. If someone rents an asset through the sharing economy then they do not have such a first claim. There is always some degree of product scarcity risk, a risk that the assets have already been claimed and are therefore not available when the consumer needs them. Perceived product scarcity risk will reduce the use of sharing economy platforms (Lamberton & Rose, 2012). sharing economy services will assign priority in different ways: with peer-to-peer services, the owner will have first call on their assets; other platforms operate on a first-come-first-served model where booking in advance gains priority; Uber encourages greater supply with surge pricing to diminish the extent of the problem; and it would be possible to auction first claim on an asset or allow someone to sell or give away their first-come-first served right to use it. Product scarcity risk is still clearly a barrier, particularly in immature or less dense markets where supply is likely to vary more.

To the extent the sharing economy extends into new markets where product scarcity risk is particularly serious (e.g. human health services), platforms might need to develop new means of assuring consumers that assets will be available when needed. Platforms might create some kind of reserve of providers who are paid to act as a provider of last resort in the event that others are unavailable, creating a capacity market.

4. Labour market obstacles

There might be a range of reasons why those who are unemployed or under-employed might be unable to take up work in the sharing economy, including:

− Low mobility. If they are unable to travel either to areas in which there is more demand for work, or are unable to travel to local places of work, then it might be difficult for them to work. − Sticky wage demands. There might be a range of reasons why workers are unable to accept the wages on offer from sharing economy platforms, even if they would otherwise be unemployed or under-employed. These could include a potential impact on the worker's entitlement to benefits; expectations of higher earnings in other potential work which they do not expect to be unable to search for while working in the sharing economy; or a pure aversion to a reduction from earlier earnings. − Technical skills mismatches. Some workers might not possess technical skills needed by sharing economy platforms, such as the ability to drive or to carry out household maintenance tasks. − Social skills mismatches. Some workers might be unable to demonstrate the reliability or standards of customer service needed to maintain an adequate rating score and thereby remain active on sharing economy platforms. This could potentially contribute to social exclusion.

PE 558.777 I - 90

5. Trust

Consumers renting goods or buying services in the sharing economy need to trust that the service will be delivered to a reasonable standard at the expected price, or that they will get proper compensation if it does not, and that their safety and security will be maintained. Those providing goods or selling services need to trust that any assets shared are not mistreated, that they will be paid as agreed and that their own safety and security will be maintained.

There are several strategies by which platforms might try to create that trust, including insurance, prior scrutiny before participants in the market start using the platform and ratings once those participants have started using the platform.

Many of the strategies which might promote trust, by preventing participants being exposed to potentially unscrupulous counterparties, will themselves constitute something of a barrier to the growth of the sharing economy. Those participants who cannot meet the standard required will be excluded and platforms cautious about consumer trust might err on the side of caution, excluding participants when in doubt. They are also expensive, raising the bar in terms of the value of a transaction or creating a barrier to entry in these markets.

5.1. Insurance Insurance is most likely to be helpful in reassuring participants that they will not be exposed to excessive risks without compensation in the event of problems that are not their fault. It will be less effective to the extent that problems are too serious for compensation to be adequate (e.g. violent crime) or that problems are too minor and difficult to substantiate for insurance to operate (e.g. poor service).

There are challenges in offering insurance to sharing economy providers (Accenture, 2015). The risks in the new markets created by sharing economy platforms are often poorly understood. Insurers might struggle to quantify those risks and the insured might not understand which risks are covered. Fixed costs will often need to be lower than for other commercial insurance given the relatively small volumes of transactions over which they can be spread, providers often find commercial insurance rates prohibitive.

Those problems have already been overcome to some extent by many platforms. Most sharing economy platforms arrange or insist upon some form of insurance for market participants. Under the 'TaskRabbit Guarantee' tasks are covered against certain risks up to $1m. BlaBlaCar has arranged with insurance firm AXA to offer free additional insurance cover to members using its service, which includes allowing passengers to share the task of driving (AXA, 2015). Uber in the UK insists that its drivers have commercial insurance. In other jurisdictions it provides additional insurance for drivers using the UberX or UberPOP service.

Over time, it might become possible to reduce insurance costs further on the basis of the data that sharing economy platforms are able to collect. First, the same ratings that enable participants to decide whether to transact with one another, either in their current form or in some improved form over time, might prove predictive of insurance risk. Second, the platform might be able to directly collect data of interest to insurers. Uber might be able to provide

PE 558.777 I - 91

journey records with information on drivers and their driving style, for example, in order to reassure insurers and thereby reduce premiums for drivers.

5.2. Prior scrutiny

Prior scrutiny is most likely to be helpful in reassuring participants that their counterparties will not be dangerous, or otherwise obviously unsuitable, on the basis of their past behaviour. It will be less effective to the extent that past behaviour, in other situations, is not felt to be indicative of future behaviour, or if there is no existing source which either measures or can effectively proxy for the problems concerned.

There are two principle sources that could be used for the prior scrutiny of market participants by sharing economy platform: public sources, such as criminal records; and private sources, such as credit records.

Public sources are already used to some extent. Uber requires drivers to submit an official check against their criminal record. The challenge in extending the use of this public source is that criminal record checks are often difficult to access, with fees to obtain the records and delays. The extent of this challenge varies by Member State. The standard criminal records check in the UK takes around two weeks and costs £26, whereas in Estonia the punishment register can be accessed online or by email at a cost of €4.

Private sources are used in similar transactions and might be integrated into sharing economy platforms over time. Credit records are already used for a range of other transactions and maintaining a high-quality credit record might be seen as indicative of broader reliability. There will be a cost in buying and using these services, but sharing economy platforms might also be able to provide data valuable to credit scoring.

5.3. Ratings

Ratings allow users to ascertain whether or not a consumer has behaved well in interacting with other market participants in the past.

There has been some discussion of the involvement of government in the devising or kitemarking of ratings systems. There are concerns that if this is done, or in particular done too early, it might stifle innovation, the targeting of reputational systems to the specific needs of consumers and providers in specific markets, and the ability of new platforms to differentiate themselves by the quality and nature of their reputational systems.

PE 558.777 I - 92

6. Tax and other policy

6.1. Tax policy

While we considered the impact of the sharing economy on tax compliance earlier, tax policy might in turn, in some cases inhibit the development of the sharing economy in two ways:

− High taxes on the returns to establishing sharing economy platforms might mean that fewer platforms are established. That might have a number of effects: reducing competition between platforms; hindering the development of potential European competitors to US platforms; and potentially slowing the development of new business models. − High or complex taxes (creating a compliance burden) might discourage providers and lead to a reduction in supply. That might reduce producer and consumer surplus. − If taxes are not 'neutral' in the sense of being applied only to some kinds of platform or service and not others, that could distort the development of the sharing economy, undermining its ability to achieve its full potential.

There are a number of studies which have found that a relatively heavy tax burden will impede new business creation and that, for example, a '10 percentage point increase in the first year effective corporate tax rate reduces business density by 1.9 firms per 100 people (average is 5), and the average entry rate by 1.4 percentage points (average is 8)' (Djankov, Ganser, McLiesh, Ramalho, & Shleifer, 2010). The effective average corporate tax rate in the EU28 is around 21 per cent (Eurostat, 2014, p. 31).

The compliance cost of taxes for providers might be reduced using the data collected by platforms. Providers might be able to generate earnings reports in a form designed to be simply entered into a tax return, for example, or they might enable the automatic filling out of tax returns, to be approved rather than created by the individual taxpayer.

It is standard in the theory of optimal taxation to observe that if some activities are taxed more than others, without good reason, that may distort economic activitiy. That could occur if tax rates are higher for some activities than others. It could also occur if taxes are enforced on some activities but not enforced on others.

6.2. Public sector provision

In some of the services that the sharing economy might enter, there is established public provision. Outpatient care could in theory be provided through sharing economy platforms, for example, and there are platforms like Pager in the United States which offer such services. In many European countries, such care is provided entirely or in part by public bodies with funding from general taxation or specific levies or social insurance schemes.

In some cases, public provision might crowd out the sharing economy. Platforms might not be able to establish themselves in markets where there are established providers for which consumers have already paid.

PE 558.777 I - 93

In other cases, however, platforms might either be able to offer distinct services, serving niches not served by public provision (e.g. providing appointments to patients who want to see a doctor particularly quickly), or to function as part of the public provision of healthcare services. Healthcare organisations might use sharing economy platforms essentially as a commissioning mechanism for certain services. While this would undoubtedly require some innovations in the platform, as certain obstacles like product scarcity risk would be likely to grow more salient, it might allow for a greater utilisation of scarce members of staff (the effects would likely not be as dramatic as with existing services such as Uber, though, as in many cases the need for demanding professional qualification would limit the potential for supply to increase).

7. Regulation

There are a range of regulatory barriers that might inhibit the development of the sharing economy (for better or worse), including:

− Outright or effective bans on sharing economy platforms. − Regulatory costs which deter self-employment. − Regulatory costs which deter marginal transactions. − Inconsistencies or idiosyncrasies in intellectual property rules.

7.1. Outright or effective bans

There are a number of Member States in which sharing economy applications have been outright banned because they are not compliant with regulatory structures applied to established providers, or are subject to regulatory requirements which most sharing economy providers are not able or expected to meet.

A judge in Spain banned the UberPop service, arguing that drivers on the platform 'lack the administrative authorisation to carry out the job, and the activity they carry out constitutes unfair competition' (BBC, 2014). By April 2015, there had been injunctions against the firm in five Member States: Belgium, France, Germany, the Netherlands and Spain (EurActiv, 2015).

Airbnb has also faced legal challenges in France (Schofield, 2014) and Spain (Kassam, 2014). The justification for both was a failure to abide by accommodation regulations. These bans can have effects that cross national borders. First, they inhibit the development of services that cross Member State borders which might thereby encourage other business between Member States (by creating more familiar services for travellers). Second, they potentially favour local providers of booking services. Second, we understand that some of the rulings prohibit consumers using the services in other Member States, where they are clearly legal (e.g. we have been informed that Spanish consumers might be prevented from using Uber in the UK).

PE 558.777 I - 94

7.2. Regulatory costs which deter self-employment

To the extent that sharing economy platforms allow people to more easily establish themselves as self-employed (and recognising that the employment status of those providing services via certain sharing economy platforms has been the subject of legal dispute), any barriers to people establishing themselves as self-employed will become obstacles to the growth of the sharing economy. The Commission Communication on a modern SME policy for growth and employment in 2005 identified a range of measures that might be taken to improve the environment for self-employment, which included cutting red tape and improving SME access to markets.

In a more recent Eurobarometer survey (2010), most respondents did not identify administrative difficulties and the burden of red tape as a reason why a business start-up would not be feasible. Lithuania and Hungary were exceptions to that general trend, however, with 22 and 17 per cent of working age respondents respectively saying that it would be unrealistic to become self-employed because of the administrative difficulties they would face. Other research has identified regulatory burdens to self-employment an area where some Member States could still take action (European Employment Observatory, 2010, p. 12):

'In several of the EEO countries, the self-employed are said to have been subject to an administrative burden — which governments are now beginning to address. In France, for instance, self-employment has been hindered by a complex administrative system and high social taxes, although this is now being helped by the introduction of the 'auto-entrepreneur' status, which relaxes burdens so that employees, students and retired people can take on self- employment activity. In Hungary, there are governmental regulations which go against setting up a business, in particular a small business, and in Croatia, administrative procedures have been long and arduous for any kind of business activity and/or enterprise.

Registering a craft or enterprise is not complex in itself, but registering and starting up a particular activity in Croatia is usually administratively demanding, tiresome and lengthy. However, administrative regulation has recently been simplified.'

One more up-to-date indicator for the difficulty in establishing as an independent contractor is contained in the World Bank Doing Business survey. The 'distance to frontier' scores34 (in which a higher score represents an greater ease in starting a business) for EU Member States vary from somewhat below to somewhat above the norm among high income OECD economies, with no immediately discernible pattern in terms of higher or lower income or newer or older Member States.

34 These scores represent a simple average across the component indicators in starting a business: number of procedures, time, cost and minimum paid-in capital. The World Bank describes the calculation of distance to frontier scores as follows: 'An economy's distance to frontier is reflected on a scale from 0 to 100, where 0 represents the lowest performance and 100 represents the frontier. For example, a score of 75 in DB 2014 means an economy was 25 percentage points away from the frontier constructed from the best performances across all economies and across time. A score of 80 in DB 2015 would indicate the economy is improving.'

PE 558.777 I - 95

Figure 8: Ease of Starting a Business, distance to frontier, data collection completed June 2014

Sources: World Bank, Doing Business, Starting a Business

It may be that if sharing economy platforms reduce other barriers to becoming self-employed (by establishing new market opportunities, or reducing the need for capital investment as people can use assets they already own) then regulatory costs may become more salient as an obstacle.

7.3. Regulatory costs which deter marginal transactions

Any regulatory costs are most likely to deter marginal transactions.

First, if someone has alternative uses for a given asset then they might be less willing to consider a welfare-improving use of that asset in the sharing economy if it entails regulatory costs. An owner of a hotel would get no value from that asset if they decided to leave it empty, whereas an owner of a flat might get some alternative use out of a spare room if they did not rent it (e.g. using it as storage). The owner of a flat might therefore be more likely to not rent it out if there is an administrative hassle involved in doing so.

Second, fixed costs of regulation will be diluted over fewer transactions. If someone is using a given room to accommodate tourists every night of the year, then it might be worthwhile for them to make install certain features required by regulation, whereas if someone is only letting a property for ten days a year that might not be worthwhile.

PE 558.777 I - 96

For these reasons, strict equality in the application of regulation might not produce equal results in practice. Regulatory burdens which might have modest effects on capacity in traditional markets might be prohibitive in the sharing economy.

7.4. Inconsistencies or idiosyncrasies in intellectual property rules

The development of platforms might be hindered by certain features of existing intellectual property rules. For example, geo-blocking (the practice of restricting access to content based upon the user's internet-determined location — which may not always, in practice, correspond to the geographical location) might mean that digital platforms and portals could not always operate effectively across borders.

PE 558.777 I - 97

Chapter 5 – What is the economic value of the barriers or obstacles?

The obstacles discussed in the last section can be considered in order, as a necessary sequence of obstacles which need to be overcome for under-utilised assets to be deployed in the sharing economy. The earlier barriers represent the technological and market context in which policy has to be formed, while the latter barriers represent the choices made within that context.

Figure 9: Obstacles process

PE 558.777 I - 98

We model each of these barriers as reducing the potential of the sharing economy in turn and develop a high-level percentage reduction over the 'shorter-term' — around the next five years, to capture obstacles likely to decline with growth in the sector — and then over the 'longer- term' — beyond that. The percentage represents the share of potential transactions which we expect might be prevented at each stage. The regulatory obstacles will be considered differently, as they depend on a policy formation process which cannot be modelled within a simple quantitative analysis of the sort we are conducting here. We will therefore test scenarios for different levels of regulatory obstacle. In either case, we will apply each percentage in turn to the €572bn value of the under-utilisation of assets estimated in Chapter 3. In this way we capture that the later, more policy-amenable, barriers will only be binding in the event that activity in the sharing economy is not prevented other, more fundamental barriers.

In some cases, it is possible to assign an indicative percentage on the basis of external data. We can base our analysis of the extent of physical barriers to sharing, for example, on the shares of the population who live in urban, intermediate and rural areas. In other cases, we assign the barrier quantitatively to one of the following qualitative categories, based on the share of transactions which that barrier is likely to prevent.35

Table 6: Barrier categories

Share of potential transactions Category prevented Small 5% Material 30% Around half 50% Most 70% Overwhelming 95%

We recognise that the precise percentages chosen here are arbitrary (e.g. 30 per cent rather than 32 percent or 25 per cent). However, we do believe that we can provide convincing arguments for the qualitative categorisations and that the numbers we ascribe to these categories are reasonable approximations likely to produce a good guide to the aggregate potential of the sharing economy when combined.

It is worth noting at this point that some very fundamental obstacles have already been included in the earlier consideration of the potential to increase the utilisation of assets. The question of whether someone will own their primary residence, for example, might be seen as a question of ownership preference (someone might rent accommodation through a sharing economy platform instead of owning a home or renting a property through the long-term rental market), but we have assumed that homes will continue to be an asset that people own or rent through conventional markets and the sharing economy market is for short-term accommodation on trips. Equally, we used data for consumption based on individual consumption, and therefore the high share for public provision in many sectors (e.g. healthcare) has already been excluded. We will attempt to avoid double counting here.

35 One can think intuitively of our five categories as being the scores 1 to 5 to a horizontal qualitative 'How much' question.

PE 558.777 I - 99

Table 7: Obstacle effects on sharing economy potential, percentage reductions

Shorter- Longer- Obstacle Justification Justification term term Over time we expect that the incentive created by the sharing economy to own a smartphone The average Member and increased penetration of the Digital State share of the Small technology more broadly will access and 55% population without a (5%) mean those without access are skills smartphone in 2013. uninterested for other reasons. This will therefore not be a barrier which is significantly binding. Sectors not amenable to Over time we expect new and sharing due to transport existing platforms to extend to costs were largely filtered intermediate areas (e.g. suburbs early in this process, so and small towns). Provision in this reflects the share of rural areas could be material with Physical the EU population living 60% 24% changes in platform business barriers outside predominantly models, but we have excluded urban areas (40.4 per cent them here (24.2 per cent live in live in predominantly predominantly rural areas urban areas according to according to the new Eurostat the new Eurostat urban- urban-rural typology). rural typology). We have already excluded sectors to reflect pure Product scarcity risk is likely to ownership preferences remain significant, but will (e.g. for some categories decline with a rise in the critical of personal effects). mass attained by platforms and Product scarcity risk is a improved management of significant barrier, demand peaks. The preference Around Consumer however, and in the Material for customisation can be half preferences shorter-term there are a (30%) expected to decline over time in (50%) number of included those assets where sharing sectors where a taste for economy penetration is greatest diversity between and increase in those sectors consumers is significant without sharing economy (e.g. cars, where penetration (e.g. mobile phones customization is and other personal effects). common). This is based on assuming While there may be changes in labour market obstacles the extent of labour market are material (30%) for obstacles, particularly with a those Member States with reduction in skills mismatches if unemployment ten per workers respond to the Labour cent or greater in 2013 incentives created by platforms market 18% 18% and small (5%) for those growing particularly strongly in obstacles with unemployment certain sectors, we do not under ten per cent in assume any reduction here as 2013. An average some features of the labour weighted by active markets concerned are quite population is then used. fundamental.

PE 558.777 I - 100

Shorter- Longer- Obstacle Justification Justification term term 72 per cent of people reported in a survey for We expect that this cause to PWC (2015) that they avoid transactions will decline could see themselves substantially over time, with a being a sharing economy reasonable framework of customer in the next two property and contract rights, but Material years. The same survey Small likely remain significant for some Trust (30%) also found providers are (5%) types of sharing (particularly if already distributed across there are high-profile bad age groups, for example, experiences). The experience of and we therefore think e-commerce payments suggests there are no demographic that, in due course, confidence is limitations significant at likely to increase significantly. this high level. Sectors subject to public provision have already While the impact of tax policy will been excluded in the decline over time, as the nature of the data sets increased prevalence of B2C used to estimate the business models means that tax Tax and Material potential of the sharing Material complexity is less of an obstacle, other policy (30%) economy. We believe tax (30%) rates themselves might not barriers policy will reduce the change and compliance might scale of the sharing rise, with the side effect of economy, however, by potentially deterring more deterring marginal activity. transactions. We test two assumptions Regulatory for the scale of the 15-30% 15-30% We test the same scenarios. barriers regulatory obstacle, 15 and 30 per cent.

The resulting impact on the scale of the sharing economy impacts can be seen below, indicating the value of different obstacles. The remaining value after each obstacle is shown, with the difference from the previous row representing the value of that obstacle.

Table 8: Value of obstacles and barriers

Shorter-term, € Longer-term, € Notional maximum potential under- 572bn 572bn utilisation Remaining value of potential under- utilisation after deducting for… Digital access and skills barriers 257bn 543bn Physical barriers 103bn 413bn Consumer preferences barriers 51bn 289bn Labour market obstacles 42bn 237bn Challenges achieving trust 30bn 225bn

PE 558.777 I - 101

Shorter-term, € Longer-term, € Tax and other policy barriers 21bn 158bn High regulatory barriers 14bn 110bn Or…

Low regulatory barriers 18bn 134bn Memo: Cost of…

High regulatory barriers 6bn 47bn Or…

Low regulatory barriers 3bn 24bn

The results above suggest that the value of regulatory barriers will rise as other obstacles become severe. Regulatory policy will therefore become more salient over time. The largest expected reduction in the value of a category of obstacle is in the field of digital access and skills, a fundamental obstacle to the development of the sharing economy but likely to decline with rising smartphone penetration across the EU. The extension of platforms from urban environments to suburbs and small towns will also allow for substantial growth, again raising the salience of physical barriers.

We have here assumed that some obstacles related to policy will remain, e.g. labour market rigidities. If those are reduced by other policy initiatives aiming to increase the flexibility of labour markets, raise educational standards, and mitigate skills mismatches, the potential of the sharing economy may rise further. In that case, the value of reducing any regulatory barriers would also rise.

PE 558.777 I - 102

Chapter 6 – How effective is the existing EU regulatory framework in promoting and overseeing such emerging new business models?

Key findings

- Other costs of non-Europe, such as the cost of gaps in the Digital Single Market, are likely to increase in salience with the growth of the sharing economy. - There are a number of existing EU initiatives relevant to the sharing economy, both measures addressing digital services in particular and those addressing broader regulatory policy. These should address a number of potential issues in the sharing economy such as price transparency. - New policy addressing remaining issues might allow regulatory objectives (such as tax collection and consumer protection) to be achieved at lower cost and address potential problems associated with the sharing economy such as new forms of social exclusion.

Research for the European Commission has argued that 'for the sharing economy in general, it would be beneficial to have specific legislation for sharing initiatives in various industries', in order to avoid a 'lack of clarity because existing legislation does not cover certain activities and transactions' or the possibility that 'legislation developed for conventional industries is wrongfully applied to markets in the sharing economy' (Dervojeda, et al., 2013, p. 14).

More recently, Elżbieta Bieńkowska – Commissioner for Internal Market, Industry, Entrepreneurship and SMEs – cited similar concerns that providers and consumers of sharing economy services might not be sure of the rules in place (Bieńkowska, 2015):

'Providers of collaborative services are not sure about which rules apply to them. Are they complying with all relevant licencing requirements, insurance obligations and safety conditions? Users face issues related to trust, safety, data privacy and insurance against certain risks. What happens in case of an accident? Are they insured in case of an accident?'

She went on to argue that:

'If we wish to make Europe a fertile ground for new business models and operators, we cannot have 28 different national regulations, or hundreds of different regional approaches – let alone thousands of separate local responses.'

Our view is that the sharing economy is and would be able to grow under the current legislative framework. It is unclear whether any lack of clarity in the current rules will lead to low standards in areas like safety or consumer protection, it seems at least as plausible that it might lead to an excessive degree of caution over matters like the screening of providers as consumers would be cautious of using platforms associated with danger and platforms would therefore be very cautious about reputational risks.

PE 558.777 I - 103

However, although the sharing economy is and would be able to grow under the current framework, as we shall see, that does not mean that the current framework will produce the best feasible results from the sharing economy in the medium-term.

1. Related cost of non-Europe estimates

There are a number of existing estimates of the cost of non-Europe which relate to the sharing economy indirectly, though none of them capture the cost of non-Europe in the sharing economy itself. The growth of the sharing economy might mean that gaps in the Digital Single Market; the Single Market for services and the Single Market in public procurement become more important.

The direct cost of the non-Europe associated with identified gaps in Digital Single Market has been estimated at €36bn to €75bn (GHK, 2014). By some margin, the largest component in that overall cost was the specific cost of being unable to access the 15 to 30 per cent savings in IT expenditures made possible by cloud computing, due to legal and security concerns that the EU might be able to address (GHK, 2014, p. 73).

The sharing economy might represent another set of digital platforms or portals which allow for very large potential savings, either in the form of lower prices in markets with fewer barriers to entry or in the form of assets which consumers no longer need to purchase and leave idle. It might therefore raise the importance of the completion of the Digital Single Market programme recently outlined by the Commission (European Commission, 2015), or shape the environment in which that programme is completed.

In many cases, services are provided on sharing economy platforms which constitute a substitute for the direct purchase of goods. Any gaps or shortcomings in the Single Market for services might therefore also become more important over time as they affect a larger share of households' total consumption. A fuller and more effective application of the Services Directive, and other measures to more deeply integrate the Single Market in services, is already estimated to have the potential to raise the long-run level of EU GDP by €338bn to €637bn (CEPS, 2014).

Finally, in many local areas, local authorities have been involved either as providers of shared assets (e.g. bicycles) or as consumers of shared economy services (to reduce the use of shared assets). The sharing economy might therefore also affect the importance of completing the Single Market in public procurement, already estimated at between €36.5bn and €66.5bn in potential annual savings (Europe Economics, 2014), as the potential arises for substantial additional savings.

While these estimates of the cost of non-Europe are all relevant to the sharing economy, we believe that there are more specific gaps in existing policy which might be remedied to fully realise the potential of the sharing economy.

PE 558.777 I - 104

2. Existing initiatives

Given its wide-ranging nature the sharing economy is affected by an equally wide range of EU policy. There are two broad relevant areas: policy which affects sharing economy platforms as digital services; and policy which affects sharing economy services through affecting the general regulation of the services which are provided through those platforms.

Relevant policy affecting sharing economy platforms as digital services includes:

− The E-Commerce Directive (2000/31/EC), which defines that information society services are subject to the law of the Member State in which the service provider is established and that Member States cannot restrict incoming services. This principle might be extended in the context of the sharing economy by making it easier for consumers to use platforms with which they are familiar in other Member States. − The Commission is currently reviewing the legal framework on the protection of personal data, aiming to modernise the legal system, strengthen individual rights and improve the clarity and coherence of the rules (DG JUST, 2015). The principle should still be that 'personal data can only be gathered legally under strict conditions, for a legitimate purpose.' This may affect certain proposals for changes to ratings systems over time. − The Digital Single Market Strategy proposed by the Commission includes proposals to construct a new regulatory framework for online platforms, in part through a new Internal Market Strategy and e-commerce framework. As a part of the strategy, the Commission also proposes to address 'a number of concerns over the growing market power of some platforms' (Commission, 2015, p. 11). We address how this might relate to sharing economy platforms in Chapter 7.

Relevant policy affecting the markets in which sharing economy providers compete includes:

− The Services Directive (2006/123/EC) aims to ensure that customers benefit from stronger rights, higher quality services and enhanced information about providers, while businesses benefit from easier establishment, easier provision of cross-border services and simplified procedures and formalities. Under the Internal Market Strategy for Goods and Services (CWP 2015), the aim is to 'deliver further integration and improve mutual recognition in key industrial and services sectors'. Providers offering their services through sharing economy platforms could clearly fit within this principle. − The Directive on Consumer Rights (2011/83/EC) regulates contracts between consumers and traders. This generally aims to strike a balance between robust consumer protections and ensuring businesses can remain competitive. Its application to sharing economy platforms should provide for price transparency with rules against hidden charges and requiring total costs are made clear. Rules against pre-ticked boxes could, however, affect opted-in benefits for sharing economy providers (an option we consider to promote social protection in Chapter 7). − The Working Time Directive (2003/88/EC) provides for limits on working time. There is the potential for Member States to derogate many limitations on working time for those with 'autonomous decision-taking powers', which has often been applied to many

PE 558.777 I - 105

self-employed workers including those offering their services through sharing economy platforms. − The Employment Information Directive (91/553/EEC), defining how workers should be told about terms and conditions; the Citizens Rights Directive (2004/38/EC), which gives workers the right to move freely and work anywhere in the EU; and other components of EU labour law create a framework in which sharing economy providers will work, though in some cases transactions will take the form of a contract between businesses (with one of the parties a self-employed contractor) rather than one between worker and employee. − The Professional Qualifications Directive (2005/36/EC) aims to facilitate the mobility of labour within the EU by allowing those qualified in one Member State to work in their profession in another Member State without repeating the qualification process. Automatic recognition in key professions is made possible by minimum training requirements, which evaluations have shown need to be updated over time to remain relevant and sufficient. This might provide a precedent for some form of common standard, though it may be difficult in professions where there is not consensus on the kind of regulation required (e.g. taxi and chauffeur services).

3. Remaining sharing economy issues

Despite the considerable existing volume of policy applicable to the sharing economy, there are significant remaining issues that require consideration.

3.1. Achieving regulatory objectives at higher cost than necessary

Existing regulation of the sharing economy has often been based on standards set for existing providers. This is intended to provide a level playing field and the question of whether a level playing field has genuinely been achieved has been the subject of considerable debate, protest and legal action.

In some Member States, those rules have been explicitly adjusted to allow small-scale activity without additional regulation. New rules in the UK, for example, allow residents to rent their homes for up to 90 days a year without permission or registration (DCLG, 2015). Similar rules exist which allow providers to rent their homes for up to 60 days in Amsterdam.

In other Member States, there have been attempts to enforce existing rules upon new providers, with services banned or limited as a consequence. In some cases, quantity-regulation of industries means that any new provision necessarily breaks rules which limit supply in sectors such as taxi services.

The background to these challenges is that it has been quite difficult for authorities to regulate many of the services where sharing economy platforms are currently growing. They often concern interactions between a large number of relatively small businesses; transactions are often conducted in cash and outside the controlled environment of a large retail store; and many consumers (particularly in large urban environments) will only transact with each

PE 558.777 I - 106

provider once with limited knowledge about the quality and reliability of the provider with whom they are doing business and often without the reassurance of a well-known brand.

Authorities have to take on more of the burden of assuring consumer safety and other regulatory compliance than would be the norm otherwise. This means significant costs both for the regulating authorities and for the regulated providers, which it may be possible to reduce with the growth of sharing economy platforms. It may be difficult for local regulators, particularly in smaller urban areas, to justify taking full advantage of this opportunity, however, as it will involve new regulations for platforms which might at first only have a relatively small presence in the market for which they are responsible.

3.2. Market segmentation and restriction

While the regulation of the online services is conducted with strong rules to ensure a Single Market, through the E-Commerce Directive, the offline goods and services offered using sharing economy platforms are the subject of varied and inconsistent Member State or local regulation. This regulation impedes the development of a Single Market in sharing economy platforms as these platforms are not able to operate in certain Member States. This means that travellers, for example, cannot use services they are familiar with when they travel and there are enormous differences in provision between Member States. While some differences in provision would be expected due to other barriers such as the penetration of smartphones, regulatory barriers mean that some platforms have been closed entirely in some jurisdictions (e.g. Uber in France). This limits competition among providers and could therefore lead to higher prices for consumers.

It also means that the size of the market for platforms in Europe is limited. This will reduce the incentive for new entrants and might therefore lead over time to sharing economy platforms gaining durable market power — discussed in more detail in Chapter 7.

3.3. Not making use of platform data

Tax collection, in particular, is a perennial problem in many of the sectors in which sharing economy platforms operate. As noted earlier, the shadow economy is estimated to account for around 20 per cent of GDP in the hotels and restaurants sector and around 15 per cent of GDP in the transport, storage and communication sector (Schneider, 2013).

This is unsurprising given that these activities are dispersed geographically, transactions are often in cash and it can be in the interests of both parties to not report transactions. This could change with sharing economy platforms where transactions are conducted electronically and reported as they happen. Member States are not making full use of the potential of the growth of sharing economy platforms and the data that they routinely collect on transactions. While any collection of data would need to be done sensitively, in order to not violate the principles underpinning data collection rules, it represents a significant opportunity.

The challenge in establishing helpful data collection is that for it to be truly effective it will require effective co-operation with sharing economy platforms that are often wary such regulatory involvement is a means or a precursor to attempting to restrict the growth of their operations. While a regulator could insist on co-operation, platforms could respond by

PE 558.777 I - 107

curtailing their operations in that jurisdiction and legislation is then likely to be clumsy and data collection impaired relative to a more constructive engagement, where the technical ability of the platforms can be used to their full extent.

There can be significant costs in restricting supply and potential gains in both employment and consumer welfare if those restrictions were removed. Quantity regulation in sectors such as taxi services has generally been found to diminish both consumer welfare and producer earnings. OECD research finds that such rules 'typically lead to an undersupply of services. In general, the value of such entry restrictions is converted into the value of licenses, rather than the earnings of a driver who rents their cab' (OECD, 2007).

Regulators are therefore in need of a commitment mechanism to open sectors to greater supply and credibly establish that information shared by platforms will not be used to constrain supply, which might be provided by robust Single Market rules.

3.4. Potential social exclusion

We have noted above that the sharing economy has the potential to encompass a significant portion of economic and social life. One issue we have observed that this might create is a danger of a new (and potentially rather comprehensive) form of social exclusion. Users of certain sharing economy platforms whose reputational ratings fall below key thresholds are excluded from the platform (for obvious reasons). Those so excluded may find it impossible to re-enter the platform to rebuild their reputation, because they cannot update their scores once they are excluded.

Such exclusion has an important disciplining role for users and also protects other users for poor conduct. Whilst those excluded have non-sharing economy alternatives or the possibility of using other platforms, being excluded is not comprehensive — i.e. it does not significantly curtail the excluded user's involvement is social and economic life. If, in the future, however, sharing economy platforms encompass a large proportion of social and economic activity (as our analysis suggests they might) and if platforms' reputational systems take account of users' reputations on other platforms, there is the risk that some users may become excluded from sharing economy activities altogether and be unable to find a way back in even if they subsequently reform their poor behaviour.

There is also some risk that users could become excluded maliciously or frivolously. Presumably platforms have fairly strong incentives to police malicious or frivolous negative reputational assessments. However, the consequence for an individual of being mis-scored might be sufficiently bad that levels of error that would be tolerable from a platform perspective (or too expensive to be worth eliminating) might not be socially desirable from a public policy perspective.

3.5. Could sharing economy platforms naturally tend to become monopolies?

A platform is, by its nature, a single venue linking buyers and sellers. A successful platform, particularly in markets with significant network effects (and such network effects are expected to be significant and common in the sharing economy), may tend towards becoming the sole (or

PE 558.777 I - 108

overwhelming majority) player in providing the marketplace for some particular sharing economy activity.

A natural concern, therefore, is that as sole (or overwhelming majority) players, sharing economy platforms might become monopolies. The normal concern with monopoly is that it leads to overcharging, under-provision and inefficient production. This could affect both consumers and providers.

3.6. Should all sharing economy service providers be employees of platforms?

It has been suggested, particularly in the United States (Economist, 2015), that one of the key features of services provided via sharing economy platforms is that the service providers would, outside the sharing economy, naturally be employees and have additional security and benefits.

Relevant regulations that might be attached to employment include:

− minimum wage and working time regulations; − responsibility for safety and other working conditions; − employer-mandated welfare provision such as sick leave, healthcare or pensions; and − the administrative element of tax.

The most difficult of these issues is where public policy uses duties imposed upon employers as a mechanism for the provision of social insurance through welfare provision. In Europe it is much less common than in the US for employer duties to be a key mechanism for healthcare provision, but that still leaves issues such as pensions or sick leave.

PE 558.777 I - 109

Chapter 7 – What additional steps should be taken at the EU level to realise its economic potential, while continuing to balance creative freedom for business with necessary regulatory protections?

Key findings

- We believe there is an opportunity to increase tax compliance and otherwise accomplish regulatory objectives at a lower cost with the data collected by platforms, but that this will likely depend on removing quantity restrictions and/or assuring platforms that such restrictions will not be introduced. - New measures are justified to support the rehabilitation of those excluded from platforms, including the potential establishment of community platforms for that purposes, but not through the regulation of still evolving ratings systems. - There should not be a general presumption that all sharing economy platforms will be dominant and any intervention should, for now at least, be based on an application of existing competition rules that allows for the dynamism of digital markets. - Labour market regulations should not be altered to specifically include sharing economy providers, who should be allowed to remain self-employed, and platforms should be allowed (and in some cases, encouraged) to develop their own means of supplying other benefits besides cash remuneration. - Other initiatives, such as those aiming to support the competitiveness of the automotive manufacturing sector, should take account of the potential development of sharing economy markets.

In our view, additional steps could be taken at the EU level to realise the potential of sharing economy platforms, while continuing to provide necessary regulatory protections. We explore options for additional steps in five areas:

− Improving market regulation to reduce costs, avoid market segmentation and restriction and improve the use of platform data to address public policy objectives. − Mitigating social exclusion from the sharing economy. − Dealing with potential market power of sharing economy platforms. − Applying labour market regulation to sharing economy platforms. − Ensuring other initiatives reflect the growth of the sharing economy, e.g. in data protection and manufacturing.

PE 558.777 I - 110

1. Improving regulations applied to sharing economy platforms

We discussed in Chapter 6 the remaining issue of not making use of platform data, which could improve tax compliance in particular but also aid decision-making more broadly.

The model in the City of Amsterdam offers an illustration of how regulation might take positive advantage of the possibilities created by the technologies underlying sharing economy platforms. The City and Airbnb have signed an agreement which, among other things, requires the firm to collect the city's tourism tax on behalf of providers. In doing so, it will increase compliance with the tax among its providers (in theory, to 100 per cent) and remove an administrative burden for those providers. In effect, the City administration has outsourced the enforcement of provider regulation to the platform. It can then verify compliance at the level of the platform itself, where the information is already centralised and can much more easily be tested.

This kind of approach, having regulatory functions undertaken by the provider, could work in a number of areas:

− Registration and identification of market participants. − Screening of market participants for criminal records and other checks where necessary. − Requiring or providing suitable insurance. − Disabled access. This might be implemented with general standards for the assets used, or by a requirement to ensure a certain level of provision for disabled access (which might be provided across a platform, or across the set of platforms subject to the regulation). − Confirming tax receipts. Information might be submitted in order to allow the authorities to check provider tax returns and platforms might also assist providers in reporting their taxes in the first place (e.g. by sending them the information that should be entered in tax forms, with a clear description of where it should be entered). − Collecting taxes. In theory, it might be possible to move steadily closer to direct real time collection of taxes and other social contributions, even if more incremental steps are taken in the near term, steadily reducing the need for self-employed providers to administer their own tax affairs.

There is a model for how this might be done which has been adopted in a number of US cities in the distinct category created for Transport Network Companies (TNCs), a term for sharing economy platforms in the transport sector which are regulated with distinct rules from traditional taxi services (PBOT, 2015). In some jurisdictions, TNCs themselves are responsible for conducting background checks. In some cases, TNCs that are unable to provide wheelchair accessible service themselves must make arrangements with another company that can provide such services. In the US, Uber reports driver earnings to the Internal Revenue Service (Steverman, 2015).

The common principle however, along with the City of Amsterdam agreement with Airbnb, is to set reasonable regulatory requirements and then look for the lowest-cost means by which the

PE 558.777 I - 111

platform can ensure providers (and consumers) meet those requirements. Our view is that such a principle could and should be extended to other sectors and the EU as a whole. The process described requires three steps.

First, define required standards. These could include requirements for the identification of providers, prior checks, reporting of provider earnings, insurance and access. However there is not in our view normally a case for quantity regulations.

Second, identify the least-cost level at which the required standards can be enforced. Where possible, the emphasis should be on regulation at the platform level, avoiding the need for enforcement at the level of the individual provider (as individual providers are likely to spread the costs of understanding and interpreting the rules over a much smaller volume of transactions).

Third, examine whether the regulation of existing providers should, in some areas, be adjusted to reflect the increased competition arising from sharing economy platforms. It may be possible to adjust regulation in some areas to reflect the increased range of options available for consumers.

Crucially, the aim in setting a level playing field should not be to subject sharing economy providers to the same set of regulations as existing providers but instead to ensure that the sharing economy meets the same underlying regulatory objectives.

1.1. Potential solutions

In order to apply this at the EU level, there are broadly three options.

1.1.1. Define a common objective

European policy could define a common objective: removal of the restrictions to the growth of the market and regulating at the platform level at the lowest-possible cost, but leave the implementation as a matter for Member State or local regulators. This might imply slower progress than common standards, and a greater degree of segmentation in the development of the market, but it would also maximise the flexibility for experimentation with different policy choices in different Member States. This might allow for the discovery of appropriate standards which might later be subject to harmonisation at the European level.

1.1.2. Define a common method

Platforms might be defined as a type of entity to which regulations can be applied and legitimate methods set out by which Member State or local regulators might set rules. This could allow competition between business models across the EU, with a common framework to which platforms could adapt, while still granting significant discretion to reflect local circumstances.

PE 558.777 I - 112

1.1.3. Define common regulations

While still allowing for some measure of local variation, European policy might set a common set of regulations for sharing economy platforms, maximising the ease with which businesses can expand across the EU and the viability of new EU entrants. This harmonised model is therefore likely to realise the greatest sharing economy potential most quickly, however it would likely mean that the regulatory standards would need to either be modest enough that compliance was realistic in all Member States or the framework flexible enough to function in all those environments.

1.2. Recommendations

Our favoured approach at this stage would be to use common regulation in a fairly narrow set of sectors (e.g. passenger transport) and regulatory issues (e.g. forbidding Member State authorities from imposing quantity restrictions), where the sharing economy is already fairly well defined and the regulatory issues are fairly clear. This would allow for the development of a Single Market in sharing economy platform services and reduce the risk of the market developing in a segmented fashion. However, that should be done with a view to rolling back such common regulations in the medium-term, so as to ensure that new sharing economy developments are (a) not curtailed or driven along pre-determined lines; and (b) can be responded to flexibly and appropriately by Member State authorities. This would reflect the principles outlined by the European Commission as a part of its Better Regulation agenda (European Commission, 2015).

2. Mitigating social exclusion

We discussed in Chapter 6 the remaining issue of potential social exclusion, arguing that the consequence for an individual of being excluded from sharing economy platforms due to poor ratings might be sufficiently bad that levels of error (either due to mistaken of malicious ratings or an inability of a market participant to rehabilitate themselves after genuine lapses) that would be tolerable from a platform perspective (or too expensive to be worth eliminating) might not be socially desirable from a public policy perspective.

2.1. Potential solutions

We shall now set out various possible means to address the social exclusion problem and their pros and cons. The reader should note that our list will include options we shall recommend against.

PE 558.777 I - 113

2.1.1. Tolerating some social exclusion

One approach would be laissez faire. We could take the view that those who conduct themselves poorly have no-one but themselves to blame if they become excluded. There would presumably be a reasonable number of people who had conducted themselves poorly in the past but who would subsequently reform. There would be fairly clear incentives for the market to provide mechanisms for such would-be reformers to re-establish their reputations (e.g. high-cost reputation-building platforms, akin to the high-cost credit lines and credit cards that are used for credit reputation repair in current financial services markets; the use of deposits when a reputation becomes low enough; or guarantor systems whereby someone with a sufficiently high reputation stood behind an excluded user), and those wanting such services could avail themselves of them. And if the market did not provide adequately for everyone, charities would provide similar mechanisms instead.

This approach would have the advantage of maximising the set of would-be reformers seeking to rebuild their reputation, which would mean there was more likely to be the critical mass required for market-based or charitable solutions to be viable.

On the other hand, there might well be some individuals that would not or could not be reached by market-based or charitable solutions and there may be no guarantee that a wide range of platforms would re-admit a rehabilitated individual even if some would. Furthermore, the social value of limiting social exclusion might be greater than the lower of the perceived value to the excluded individual (excluded users might adopt a 'who cares' façade in response to exclusion) or the amount an excluded individual might be able to pay (an excluded individual might be relatively low-income and, qua excluded individual, might find access to credit rather restricted).

2.1.2. Mandating a right to a reputational Year Zero

One wide-ranging approach to excluded users might be to grant users a right to have their reputational scores forgotten once they are excluded — so the minimum reputation score becomes that of a new user. A slightly more limited variant of that might be to restrict exclusion to the platform on which it initially occurs — so the excluding platform can exclude a poorly- behaving individual, but the worst consequence that can have on other platforms is for the individual to become like a new entrant (platforms cannot exclude a user on the basis of poor behaviour that does not occur on their own platform). This can be seen as the sharing economy equivalent of having a bad reputation in one town, so moving elsewhere to make a fresh start.

A significant downside of this approach can be seen from the traditional world fresh start: those that have a pattern of poor behaviour sometimes (perhaps often) use the opportunity of a fresh start to exploit new people that do not know of their bad behaviour. That users have achieved such a poor reputation that they become excluded from a platform is potentially relevant information for the protection of other users and the integrity of other platforms. A right to a reputational Year Zero or other restrictions upon the use of reputation data could potentially create avoidable detriment for other users and other platforms.

PE 558.777 I - 114

2.1.3. Regulating reputational scoring so that only socially desirable exclusions occur

A somewhat less wide-ranging (but still significantly interventionist) approach might be to regulate reputational scoring so that users could only become excluded in extreme cases where such exclusion was socially desirable. The point here is that the social cost of an individual being excluded from platforms in general might exceed the cost to the individual and the platforms. That might mean that some comprehensive exclusions that would be desirable from the point of view of the platforms might be undesirable from the social point of view. The solution in this case would not be to forbid exclusion. Instead, it would be to use regulation (or some other intervention such as a kitemark) to attempt to restrict exclusion to the socially desirable cases.

A downside of this approach is that it would be likely to imply regulatory authorities (or those administering a kite mark scheme) having quite extensive oversight and control of reputation scoring mechanisms. That risks impeding reputational scoring innovation and creating barriers to entry for new players. It also risks forcing reputational scoring into a common mould, despite the fact that platforms themselves may not consider reputation changes in some dimensions relevant to them — e.g. someone that frequently failed to keep their rooms adequately clean for a room-sharing platform might nonetheless be reliable in turning up on time and performing work properly on a task-sharing platform.

2.1.4. Creating community platforms where reputation can be rebuilt

Amongst the concerns for the laissez faire approach were that some individuals might not be reached by market-based or charitable solutions and that some important platforms might not accept rehabilitated individuals. One potential solution to this would be for the state to create specially tailored community platforms where individuals can rebuild their reputations. Natural versions of this might include community services tasks or relatively simple work tasks. We can think of this as the sharing economy version (albeit on a less dramatic scale) of the ways that re-integration services arrange jobs for ex-prisoners.

One risk with this approach is that sharing economy platforms might not accept the reputation scores built up on such platforms. Indeed, there might even be a fear of stigma — that those resorting to using such a community platform might experience a downgrade in their private sector reputation. To address this, the relevant state authorities might broker agreement from socially or economically significant sharing economy platforms to take account of community rehabilitation platforms, perhaps even offering subsidies or deposits as guarantees.

2.2. Recommendations

In our view, the level of intervention required to prevent any social exclusion would be prejudicial to trust on the part of other market participants and therefore the development of the sharing economy. As we have seen in Section 4.5, trust is one of the key potential obstacles to the development of the sharing economy and many measures platforms might naturally take to build and preserve trust would be likely to entail some social exclusion. Therefore there will

PE 558.777 I - 115

need to be some tolerance of social exclusion and therefore the right to a reputational year zero appears to us to be undesirable.

The tolerating of some social exclusion is likely to secure higher social acceptance if there are seen to be state initiatives that provide a route out of social exclusion. The least intrusive form of such state initiatives would be the establishment of community rehabilitation platforms to allow the rebuilding of sharing economy reputations. Therefore we recommend these.

Whilst there could be a case in the long-term for mandating fairness in platform exclusion (akin to policies such as treating customers fairly in financial services regulation), in our view sharing economy reputation scoring systems would need to be much better developed before an appropriate regulatory standard for fairness could be defined and enforced. A kitemark could be an alternative if it were robust, but a kitemark based upon the current as-yet limited understanding of ideal reputation scoring would lose credibility and in the meantime might mislead those customers that it re-assured. We therefore recommend against either version (mandated fairness or kitemarking) of this option at this stage.

3. Dealing with potential market power of sharing economy platforms

We discussed in Chapter 6 the remaining issue of addressing concerns that sharing economy platforms might become monopolies. We noted that a successful platform might create network effects that would allow them to become the sole (over overwhelming majority) player in providing the marketplace for some particular sharing economy activity.

3.1. Potential solutions

Relying upon market forces and innovation to undermine market power

A platform is a two-sided market, linking consumers and providers. There is a well-known case that in such markets even a sole dominant provider prices such that social surplus is maximised. (Gonçalves, 2003) However, policy practice has been, nonetheless, to regard dominant networks as giving rise to a potential competition issue. sharing economy platforms have emerged (and by their nature operate) in a technically dynamic sector in which contestability is high — i.e. even a sole provider might quickly be replaced by another sole provider. Unlike utility networks, the need for capital investment is relatively low and within the reach of a wide range of potential entrants. Furthermore, although the first entrant in a sharing economy market may face costs in establishing trust, consumer understanding of the product, and a stable market structure, it is likely that for later entrants that key spade-work would already have been done, considerably lowering the costs of follow- on entry.

As well as potential start-ups, new platforms may be created (indeed, are already being created) by firms in other sectors: manufacturers offering their products directly to market (e.g. BMW

PE 558.777 I - 116

creating a car-sharing platform); technology firms integrating new platforms into existing service offerings (e.g. Apple Music created as a rival to Spotify); or even providers themselves establishing platforms with which they are comfortable.

Therefore there might be scope, even when a sharing economy platform has a high market share, to rely upon market forces to restrict market power.

3.1.1. Developing the Single Market so as to maximise the size of the market, creating the greatest scope for multiple platforms

Providers are more likely to find customers more quickly in larger platforms, but with a sizeable market any power is likely to be limited. Despite the network effects, with the development of a Single Market in Europe there might be room for multiple platforms to reach critical mass, based on different offerings which appeal to different consumers and/or providers.

This might mean that the development of the Single Market could, in itself, be a solution to potential market power.

3.1.2. Relying upon existing competition rules to identify instances of market power and specific appropriate interventions

If market forces and the development of the Single Market failed to prevent a sharing economy platform from securing durable market power, normal competition rules in the EU would identify that and mandate appropriate policy interventions.

This might mean that no sharing-economy-specific rules to address market power are required. In the proceedings of a recent workshop for the ITRE Committee it was argued, however, that competition policy 'should treat digital markets carefully and give priority to dynamics of competition', 'recognising dynamism in markets' (Van Gorp, 2015). This might include both actual dynamism (observed entry and exit) and potential dynamism (the potential for firms to enter and exit). Given the market forces and potential competitors identified earlier, this implies a cautious approach to the exercise of competition powers in this area.

3.1.3. Treating sharing economy platforms in a manner analogous to regulated utilities

In a number of sectors where provision tends to monopoly — e.g. energy, water, telecommunications — the EU has established a norm for economic regulation. Similarly, certain platforms in other contexts (e.g. broadcasting) have been deemed to have market power and subject to ex ante or ex post restrictions. If sharing economy platforms had sufficiently common monopoly features, one option might be to set a similar framework for the economic regulation of sharing economy monopolists. Whether such a framework was appropriate and whether any specific sharing economy platform should be covered by it would of course need to take account of the technically dynamic nature of digital platforms and the extent to which,

PE 558.777 I - 117

even if the market would tend to produce only one platform at a time, that platform might be under continuous significant threat of being replaced by a new entrant (i.e. the market being contestable).

The best form of such economic regulation (if economic regulation were indeed appropriate) would depend upon the specific details of the market involved. Some economic regulation involves the setting of price or revenue caps (e.g. that is standard in sectors such as telecoms or water); other economic regulation involves setting a permitted rate of return (e.g. that was the system used under the Pharmaceutical Price Regulation Scheme in the UK). In other contexts economic regulation takes a more general ex post form of checking, after the event, that the achieved rate of return achieved was not excessive.

3.2. Recommendations

In our view, in at least a significant range of cases within the wide sharing economy, some combination of competition and contestability, potentially fostered and facilitated by the extension of the Single Market, should be adequate to curtail market power. There should not be a general presumption that all sharing economy platforms will be dominant.

It falls outside our scope here to offer a judgement on whether (and if so which) specific sharing economy markets should be deemed to have significant market power and be subject to remedies. However, to the extent that competition, contestability and the extension of the Single Market do not undermine market power, we believe that competition authorities are the most appropriate next step, before recourse to general mandating of economic regulation.

4. Applying labour market regulation to sharing economy platforms

We discussed in Chapter 6 the remaining issue of whether providers should be considered employees of platforms, noting that there are a number of relevant regulations that might be attached to employment, including:

PE 558.777 I - 118

4.1. Potential solutions

4.1.1. Mandating that all sharing economy service providers be platform employees

Under this option the sharing economy platform would have responsibility for the working conditions, meeting of minimum pay requirements, and the provision of welfare benefits such as pension contributions that apply to corporate employers.

4.1.2. Creation of a new employment status of 'sharing economy service provider'

Under this option sharing economy platforms would not be subject to all of the responsibilities of employers, but would not be able to treat users as fully-separate businesses. Instead, there would be a more limited set of duties that would apply, e.g. dealing with taxes or closing the platform to those seeking greatly excessive working hours. This latter could be seen as akin to the duties some countries impose upon casinos to exclude users who lose more than a certain amount in a given period.

4.1.3. Avoiding extensive roles for employers in public welfare provision

In an economy dominated by traditional employment, it might be feasible to use employers as the means to fund a wide range of social benefits such as sickness and health insurance or pensions. In an economy in which self-employment was much more common, such an approach would be much less feasible. As part of the wider movement towards self- employment, the development of the sharing economy might mean that policymakers should in the future avoid extensive roles for employers in public welfare provision.

4.1.4. Encourage or facilitate platforms in developing their own user benefits

Under this option it would be hoped (or sought) that platforms would develop their own additional user benefits. It is likely that platforms will develop in which users of a platform are offered an option to have automatic deductions from their earnings on the platform to pay for an insurance policy against sickness. Policy might either seek to make such options universal or to have a default 'opted in' position akin to recent pension reforms making opting in to employer pension schemes the default.

4.1.5. Extension of insurance and other financial markets

The self-employed already have a number of options available to them to self-insure, as do employees that would prefer a higher level of protection than is available via the state or their employer. The growth of the sharing economy may mean a larger market for such products and

PE 558.777 I - 119

therefore reward innovation by insurance companies and other financial institution in providing self-insurance products.

4.2. Recommendations

As we have noted earlier, sharing economy service providers can be seen as part of a more general movement towards self-employment (indeed, the sharing economy itself can be seen as a response to that trend). It might be in the nature of many sharing economy services (e.g. assembling flat-pack furniture) that they would not and could not be offered, at reasonable cost, by traditional corporates with significant numbers of employees.

One consequence of this is that it be most natural to treat (and, if appropriate, develop) the regulation of sharing economy service providers as part of the general regulation of self- employment, rather than either mandating that all sharing economy service providers be treated as employees or by creating a new employment status of 'sharing economy service provider'.

Where sharing economy platform users have an additional main employment, that might still be the source of important employer-mandated welfare provision. If employees want the flexibility of self-employment, preventing them from obtaining that via digital platforms or portals might risk chasing such activities into the 'black economy'.

If users would prefer to be employees (and if the sharing economy becomes a larger part of the overall economy, it is possible that many users might be in this situation), it is likely that there would be an increased role for 'hybrid' platforms whereby the platform offered business the option to provide services and the business themselves had employees.

In our view, the best approach would be to allow (and in some cases potentially encourage) platforms to develop their own benefits options that would compete with the insurance products users could obtain for themselves. We do not believe there is a rationale for additional labour market regulation of platforms at this stage.

5. Ensuring other initiatives reflect the growth of the sharing economy

Apart from these areas directly related to the sharing economy, there are a number of areas in which other EU policy might be adapted to maximise the potential of the sharing economy:

− Data protection rules should retain the standard that 'personal data can only be gathered legally under strict conditions, for a legitimate purpose'. Principles such as the right to be forgotten or consumers owning their own data might need to be interpreted cautiously. It is important, for example, that platforms are able to retain information about consumers in order to enable a functioning ratings system. − Existing efforts to create a framework for growth in manufacturing sectors, e.g. the CARS 2020 Action Plan in the automotive sector, should take account of the potential for those sectors to be affected by the sharing economy. The balance of opportunities

PE 558.777 I - 120

might change in European and global markets, for example with a shift towards the consumption of fewer vehicles but a higher for each vehicle. European policymakers could encourage greater engagement between manufacturers and sharing economy platforms. − Planning in other areas — e.g. transport infrastructure — should take account of the possibility that sharing economy developments overturn existing projections for volume and need. For example, if some of the higher estimates of the numbers of cars removed from the road were to be realised, that could lead to a significant reduction in parking requirements in cities. − The ongoing reform of intellectual property rules might need to take account of the increased prevalence of a cross-border hiring model which might render geo-blocking and related restrictions otiose.

PE 558.777 I - 121

Bibliography

Accenture. (2015). Lead the Pack or Follow the Leader: Insuring Risk in the sharing economy. Arthur, C. (2014, April 30). The death of the featurephone in the UK - and what's next. The Guardian. AXA. (2015, May 5). BlaBlaCar and AXA launch first-of-its-kind ridesharing insurance product. Retrieved June 25, 2015, from AXA: http://www.axa.com/en/news/2015/blablacar.aspx Baker, D. (2014, May 27). Don't buy the 'sharing economy' hype: Airbnb and Uber are facilitating rip-offs. The Guardian. Bassanini, A., & Brunello, G. (2010). Barriers to Entry, Deregulation and Workplace Training. Munich: CESifo. BBC. (2014, December 9). Uber taxi app suspended in Spain. BBC News. Bieńkowska, E. (2015). Speech: Opportunities for innovative business models in the EU Single Market. Brussels: European Commission. Block, J. H., & Wagner, M. (2010, April). Necessity an Opportunity Entrepreneurs in Germany: Characteristics and Earnings Differentials. Schmalenbach Business Review, pp. 154-174. Bonamy, J., Dorso, A., & Guerrab, A. E. (2014). Un Taxi pour L'avenir des Emplois pour la France. Paris: Assemblée nationale. Botsman, R. (2013, November 21). The sharing economy Lacks A Shared Definition. Co.Exist. Broda, C., & Weinstein, D. (2004). Globalization and the Gains from Variety. New York City, NY: Federal Reserve Bank of New York. Brown, K. V. (2015, June 19). How much would it cost Uber to make drivers employees? Fusion. Brynjolfsson, E., & McAffee, A. (2014). The Second Machine Age: Work, Progress and Prosperity in a Time of Brilliant Technologies. W. W. Norton & Company: New York, NY. BVRLA. (n.d.). Why choose car rental? Retrieved July 20, 2015, from BVRLA: http://www.bvrla.co.uk/advice/guidance/why-choose-car-rental Cellan-Jones, R. (2015). Will London's mayor put the brakes on Uber? London: BBC. CEPS. (2014). The Cost of Non-Europe in the Single Market - II - The Single Market for Services. Brussels: European Parliamentary Research Service. Choudhary, V. (2007). Software as a Service: Implications for Investment in Software Development. Proceedings of the 40th Hawaii International Conference on System Sciences. Irvine, CA: The Paul Merage School of Business, University of California, Irvine. Commission. (2015). Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Brussels: European Commission. Cowen, T. (2011). The Great Stagnation: How America Ate All The Low-Hanging Fruit of Modern History, Got Sick, and Will (Eventually) Feel Better. Dutton. Damodaran, A. (2014, June 18). Uber Isn't Worth $17 Billion. FiveThirtyEight. DCLG. (2015). Measures to boost sharing economy in London. London: Department for Communities and Local Government.

PE 558.777 I - 122

Dervojeda, K., Verzijl, D., Nagtegaal, F., Lengton, M., Rouwmaat, E., Monfardini, E., & Frideres, L. (2013). The sharing economy: Accessibility Based Business Models for Peer-to-Peer Markets. Brussels: European Union Business Innovation Observatory. DG ENTR. (2014). CARS 2020 Report on the state of play of the outcome of the work of the High Level Group. Brussels: European Commission, Directorate-General for Enterprise and Industry. DG JUST. (2015, July 22). Protection of personal data. Retrieved July 29, 2015, from European Commission, DG Justice: http://ec.europa.eu/justice/data-protection/ Djankov, S., Ganser, T., McLiesh, C., Ramalho, R., & Shleifer, A. (2010, July). The Effect of Corporate Taxes on Investment and Entrepreneurship. American Economic Journal: Macroeconomics, pp. 31-64. Eckhardt, G. M., & Bardhi, F. (2015, January 28). The sharing economy Isn't About Sharing at All. Harvard Business Review. Economist. (2015, July 4). Schumpeter: McJobs and UberJobs. The Economist. Edelman, B., & Luca, M. (2014). Digital Discrimination: The Case of Airbnb.com. Cambridge, MA: Harvard Business School. EurActiv. (2015, April 2). Uber files complaints against German and Spanish bans. EurActiv.com. Eurobarometer. (2010). Entrepreneurship in the EU and beyond. Brussels: European Commission. Europe Economics. (2014). Cost of Non-Europe in the Single Market - IV- Public Procurement and Concessions. Brussels: European Parliamentary Research Service. European Commission. (2015). A Digital Single Market for Europe: Commission sets out 16 initiatives to make it happen. Brussels. European Commission. (2015). Tool #1: Principles of Better Regulation. Brussels. European Employment Observatory. (2010). European Employment Observatory Review: Self- employment in Europe. Brussels: European Commission. Eurostat. (2014). Taxation trends in the European Union. Brussels: European Commission. Fraiberger, S., & Sundararajan, A. (2015). Peer-to-Peer Rental Markets in the sharing economy. New York, NY: NYU Stern School of Business. Frenken, K., Meelen, T., Arets, M., & van de Glind, P. (2015, May 20). Smarter regulation for the sharing economy. The Guardian. GEWOS. (2014). Airbnb and the Berlin housing market: The impact of Airbnb properties on the housing supply in Berlin. Berlin: Airbnb. GHK. (2014). The Cost of Non-Europe in the Single Market - III - Digital Single Market. Brussels: European Parliamentary Research Service . Gilmore, J. H., & Pine, B. J. (1997, January-February). The Four Faces of Mass Customization. Harvard Business Review. Gurley, B. (2014, July 11). How to Miss By a Mile: An Alternative Look at Uber's Potential Market Size. Above the Crowd. Hall, J., & Krueger, A. (2015). An Analysis of the Labor Market for Uber's Driver-Partners in the United States. San Francisco, CA: Uber. IBM. (2011). IBM Global Parking Survey: Drivers Share Worldwide Parking Woes. Armonk, NY: IBM.

PE 558.777 I - 123

Ikkala, T., & Lampinen, A. (2014). Defining the price of hospitality: networked hospitality exchange via Airbnb. Proceedings of the companion publication of the 17th ACM conference on Computer supported cooperative work & social computing (pp. 173-176). New York, NY: ACM. Investopedia. (n.d.). sharing economy. Retrieved June 9, 2015, from Investopedia: http://www.investopedia.com/terms/s/sharing-economy.asp ITDP. (2013). The Bike-Share Planning Guide. New York, NY: Institute for Transportation & Development Policy. Kassam, A. (2014, July 7). Airbnb fined €30,000 for illegal tourist lets in Barcelona. The Guardian. Lamberton, C. P., & Rose, R. L. (2012). When is Ours better than Mine? A Framework for Understanding and Altering Participation in Commercial Sharing Systems. Journal of Marketing. Martin, E., & Shaheen, S. A. (2011, Spring). The Impact of Carsharing on Household Vehicle Ownership. Access, pp. 22-27. Mason, P. (2015). PostCapitalism: A Guide to our Future. London: Allen Lane. Meelen, T., & Frenken, K. (2014, January 14). Stop Saying Uber Is Part Of The sharing economy. Co.Exist. Mulshine, M. (2015, June 18). After a disappointing Airbnb stay, I realized there's a major flaw in the review system. Business Insider. Munger, M. (2014, July 7). Michael Munger on the sharing economy. (R. Roberts, Interviewer) Nordhaus, W. D. (2004). Schumpeterian Profits in the American Economy: Theory and Measurement. Cambridge, MA: NBER. OECD. (2007). Taxi Services: Competition and Regulation. Paris: OECD. Orsi, J. (2013, September 16). The sharing economy Just Got Real. Shareable. PBOT. (2015). Transportation Network Companies: Regulations Review & Report. Portland, OR: Portland Bureau of Transportation. PolicyInteractive. (2014). Analysis Report: New American Dream Survey 2014. Charlottesville, VA: The Center for a New American Dream. PWC. (2014). How did we develop our sharing economy revenue projections? PriceWaterhouseCoopers. PWC. (2015). Consumer Intelligence Series: The sharing economy. PriceWaterhouseCoopers. Reich, R. (2015, February 2). The Share-the-Scraps Economy. RobertReich.Org. Retrieved from Robert Reci. Rubin, A. J., & Scott, M. (2015, June 25). Clashes Erupt Across France as Taxi Drivers Protest Uber. New York Times. Samaan, R. (2015). Airbnb, rising rent, and the housing crisis in Los Angeles. Los Angeles, CA: Los Angeles Alliance for the New Economy. Schneider, F. (2013). The Shadow Economy in Europe 2013. Visa Europe. Schofield, H. (2014, December 26). Short-let apartments spark Paris row as Airbnb thrives. BBC News. Shaheen, S. A., Cohen, A. P., & Chung, M. S. (2009). North American Carsharing: 10-Year Retrospective. Washington, DC: Transportation Research Board. Slagen, D. (2014, September 8). From P2P to B2B: The next phase of the sharing economy. VentureBeat.

PE 558.777 I - 124

Smart, R., Rowe, B., Hawken, A., Kleiman, M., Mladenovic, N., Gehred, P., & Manning, C. (2015). Faster and Cheaper: How Ride-Sourcing Fills a Gap in Low-Income Los Angeles Neighborhoods. San Francisco: Uber. Sperling, G. (2015). How Airbnb Combats Middle Class Income Stagnation. San Francisco, CA: Airbnb. Stefansdotter, A., Utfall Danielsson, C. v., Nielsen, C. K., & Sunesen, E. R. (2015). Economic benefits of peer-to-peer transport services. Copenhagen: Uber. Steverman, B. (2015, February 19). Tax Advice for Uber Drivers and Etsy Sellers (This Above All: Deduct!). BloombergBusiness. Sullivan, M. (2015, July 14). Tax Challenges For The Uber Economy. Forbes. Sundararajan, A. (2015, June 22). A safety net fit for the sharing economy. Financial Times. The People Who Share. (n.d.). About. Retrieved June 15, 2015, from The People Who Share: http://www.thepeoplewhoshare.com/about/ Van Gorp, N. (2015). Cross-Competition Among Information (Digital) Platforms: Proceedings of the Workshop. Brussels: European Parliament: Directorate General for Internal Policies. Vanhee, J. (2011). momo Car-Sharing: More options for energy efficient mobility through Car- Sharing. Brussels: Intelligent Energy Europe. Wallsten, S. (2015, July 9). Has Uber Forced Taxi Drivers to Step Up Their Game? The Atlantic. Williamson, O. E. (1975). Markets and Hierarchies. New York: Free Press. Wosskow, D. (2014). Unlocking the sharing economy: An independent review. London: Department for Business, Innovation and Skills. Zervas, G., Proserpio, D., & Byers, J. W. (2015). The Rise of the sharing economy: Estimating the Impact of Airbnb on the Hotel Industry. Boston, MA: Boston University. Zhuo, T. (2015, March 25). Airbnb and Uber Are Just the Beginning. What's Next for the sharing economy. Entrepreneur.

PE 558.777 I - 125

PE 558.777 I - 126

Annex II

Cost of non-Europe in the sharing economy: legal aspects

Briefing Paper by European Institute of Public Administration

PE 558.777 II - 127

AUTHORS This study has been written by Gracia Vara Arribas of the European Institute of Public Administration (Barcelona), at the request of the Impact Assessment Unit of the Directorate for Impact Assessment and European Added Value, within the Directorate General for Parliamentary Research Services (DG EPRS) of the General Secretariat of the European Parliament.

RESEARCH TEAM Gracia Vara Arribas – Project leader Anthony De Bondt – Researcher Bettina Steible – Researcher

RESPONSIBLE ADMINISTRATOR Pierre Goudin, European Added Value Unit To contact the Unit, please e-mail [email protected]

LINGUISTIC VERSIONS Original: EN

This document is available on the internet at: http://www.europarl.europa.eu/thinktank

Manuscript completed in January 2016. Brussels © European Union, 2016.

PE 558.777 II - 128

Table of Contents

List of abbreviations ...... 128

Executive Summary ...... 129

1. Introduction ...... 131

1.1. Defining the sharing economy ...... 131 1.2. Research aims ...... 136 1.3. Methodology ...... 136

2. National legal framework: case studies ...... 139

2.1. Transportation...... 139 2.1.1. Benefits ...... 141 2.1.2. Complaints ...... 141 2.1.3. Case analysis ...... 141

2.2. Accommodation ...... 150 2.2.1. Benefits ...... 150 2.2.2. Complaints ...... 151 2.2.3. Case analysis ...... 151

2.3. Professional services ...... 159 2.3.1. Benefits ...... 159 2.3.2. Complaints ...... 159 2.3.3. Trends ...... 160 2.3.4. Case analysis ...... 160

2.4. Comparative analysis of national legal responses ...... 162 2.4.1. Complete banning and equalising to traditional services ...... 162 2.4.2. Imposing more rules ...... 163 2.4.3. Simplifying ...... 164 2.4.4. Best practices ...... 164

3. The EU regulatory framework ...... 167

4. Remaining legal gaps and issues ...... 173

4.1. Legal gaps ...... 173 4.1.1. About the definition of the sharing economy ...... 173 4.1.2. About the notion of service ...... 174 4.1.3. About the obligations of the SE platforms ...... 176 4.1.4. Regulatory framework applicable to users ...... 177

4.2. Obstacles and barriers preventing the sharing economy from reaching its full potential ...... 178

PE 558.777 129

4.3. The potential of self-regulation in the SE ...... 180

4.3.1. The concept of self-regulation ...... 180 4.3.2. Self-regulation in the sharing economy ...... 181

5. Conclusions and recommendations ...... 183

5.1. The platform level: towards a clear categorisation of platforms ...... 183 5.2. Action at the service provider level: ensuring compliance ...... 186 5.3. Action at the consumer level: creating trust ...... 188

References ...... 189

Annex I ...... 195 Annex II...... 197

PE 558.777 II - 130

List of abbreviations

B2B Business to business CJEU Court of Justice of the European Union CoR Committee of the Regions EP European Parliament EU European Union IP Internet Protocol MS Member States SE sharing economy P2P Peer to peer TFEU Treaty on the Functioning of the EU TEU Treaty of the EU

PE 558.777 II - 128

Executive Summary

The sharing economy (SE) is a business model based on sharing underutilised assets using digital platforms, either for monetary or non-monetary benefits. Many are the terms and definitions that appear to describe the phenomenon, and in this report we use the concept in the broad sense to cover most or all the activities that are generally thought of when speaking of the SE: collaborative economy, peer economy, etc.

By now, everyone agrees that the SE is shaking grounds and disrupting the market, providing a real challenge to traditional market players, authorities and consumers. In fact, the exchanges the SE facilitates are not new in themselves: people have been renting out rooms, sharing car rides, or providing cleaning services for ages, but the novelty now is the reduction in transaction costs, and the possibility platforms offers to reach consumers worldwide through one single click. Better and more sustainable use of underutilised resources (cars, houses, skills) is claimed to be a pivotal benefit SE can provide to society. Creation of employment, prevention of the shadow economy and access to new services for consumers are amounting advantages according to the SE advocates. But many are the complaints against these business models: people are renting their apartments without complying with the existing regulations for hotel accommodation; drivers are making rides for a price without complying with the taxi regulations. The industry has mobilised and several cases have been brought to court pleading the prohibition of some of the business models.

Uber for example, is the business model most challenged in Europe, and has had to stop some of its activities in different Member States (MS). Through its platform, Uber allows consumers with smartphones to submit a trip request which is then routed to Uber drivers, who – using their own cars – will provide the transport service for the price fixed by the platform. Taxi drivers claim unfair competition, consumers praise good quality at low price, and the authorities defend the need to protect the taxi sector but want to support innovation. In the meantime, the judiciary is wondering how to qualify the services provided by the platform: is it a transport company or an IP (Internet Protocol) enabled service? Several judges have submitted a preliminary ruling to the Court of Justice of the European Union (CJEU), and different new laws have been adopted so far by different MS. The European Institutions are closely following these events, and considering the need for action in view of the clear European dimension of the SE. Whether to enact new regulation, promote self-regulation and/or defend a non-regulation approach are the issues at stake. The Commission has already announced that it will provide guidance on how existing EU law applies to the SE rather that strictly regulating it.

In this report we analyse the legal obstacles/barriers that prevent the SE from reaching its full potential and the effectiveness of the existing EU regulatory framework to oversee these new business models.

To do so, we have analysed the practices in six Member States (BE, DE, ES, FR, NL and UK) and in some cities of USA around three economic sectors: transportation, accommodation and the provision of professional services.

We have detected that the transport platforms providing transport services similar to taxis (Uber and assimilated) have been the most problematic and finally banned in four of the six MS.

PE 558.777 II - 129

These platforms make three transactions between the service provider and the consumer (information, for profit payment and control over the service provision). On the contrary, transport platforms facilitating car sharing but being not-for profit (Blablacar), have been allowed provided that users comply with the basic existing local/national legislation (taxes, insurance, consumer protection, data protection, etc.). This second type of platform provides only two transactions (information and payment –not for profit) without controlling the provision of the service itself: the driver shares their ride to their destination and fixes the price.

In the case of accommodation platforms (renting houses or rooms through a platform), legal acts have been approved or adapted in different cities/countries, to oblige the platforms and the service providers to comply with different sets of rules, and to try to establish a dividing line between what constitutes a sharing practice or a professional practice illegally competing with the professional sector. In the case of platforms for the provision of services (cleaning, consultancy, etc.) legal problems have been debated in American courts so far, and mostly related to the qualifications of service providers as employees or as self-employed – with the consequent different legal treatment -, a debate that is common to other transport and accommodation platforms performing in Europe.

Since the European Union has the competence to protect the Single market freedoms, it needs to find the balance between creative freedom required by SE business and the need for regulation of the sector. Two types of approaches can be applied: (1) government control or top-down government regulation or (2) bottom-up regulation or self-regulation through reputation. Best practices deployed at MS level indicate that a mixture of both approaches will be needed. We recommend as a potential line of action:

1. The need to distinguish between the different types of platforms and the professional and non-professional divide – categorising at EU level would clear up the confusing landscape. 2. Harmonisation of reputational rating systems: the establishment of guidelines at EU level would provide a unified framework. 3. The establishment of enforcement mechanisms of compliance with the existing European rules on consumer protection, insurance, labour conditions, tax collection etc. For this purpose the partnership between authorities and platforms could support the externalisation of the control mechanisms to the platform where all the information is centralised. 4. Promote the use of self-regulation per sector based on detected best practices such as: provision of insurance contracts, performing tax collection, having in place a cash less payment system and policies to safeguard consumer protection.

In the year 2016 the CJEU will decide on the pending preliminary rulings. The European Commission will present its line for action, and different commissioned studies by different bodies will see the light. Further research will be needed to decide at EU level the precise actions that need to be taken in the course of the coming months, but we hope to have shed some light on the blurred picture of the legal aspects of the SE in Europe.

PE 558.777 II - 130

1. Introduction

The Directorate-General for Parliamentary Research Services has commissioned the European Institute of Public Administration (EIPA-Barcelona) in the person of Gracia Vara Arribas (EIPA Expert) to draft a briefing paper on 'Cost of non-Europe in the sharing economy: legal aspects'. The paper should complete the report made by Europe Economics on 'Cost of Non-Europe in the sharing economy: Opportunities and challenges' by focusing more on the regulatory aspects of the sharing economy (SE) and deepening the analysis on them.

1.1. Defining the sharing economy

When looking at the literature, a plethora of definitions appear to exist to describe what is commonly seen as part of the sharing economy (hereafter, 'SE'). Rooselaer (2014: 13) provides a non-exhaustive list of the many different concepts that have seen the light over the course of the past years when authors tried to name the phenomena. Botsman (2013), co-author of the book 'What's mine is yours. How collaborative consumption is changing the way we live', tried to provide structure to the chaos. Doing so, she identified four concepts that cover most or all activities that are generally thought of when speaking of the sharing economy.

1. Collaborative consumption: an economic model based on sharing, swapping, trading, or renting products and services, enabling access over ownership 2. Collaborative economy: an economy built on distributed networks of connected individuals and communities (as opposed to centralised institutions). Collaborative economy spans collaborative finance, collaborative education, collaborative production, and collaborative consumption. 3. Peer economy: person-to-person marketplaces that facilitate the sharing and direct trade of assets built on peer trust' 4. Sharing economy: an economic model based on sharing underutilized assets from spaces to skills to stuff for monetary or non-monetary benefits.

Building on this logic, 'collaborative economy' is the umbrella concept for both collaborative consumption and peer economy, while the sharing economy is seen as a subset of collaborative consumption.

In this Report we will follow the definition selected by Europe Economics (2015: 6)

PE 558.777 131

The use of digital platforms to reduce the scale for viable hiring transactions or viable participation in consumer hiring markets (i.e. 'sharing' in the sense of hiring an asset) and thereby reduce the extent to which assets are under-utilised.

This definition reflects the broader categories described above – including the sharing economy – such as the Collaborative Economy (Botsman, 2013), or includes sectors which others have defined as similar to but not a part of the sharing economy, those transactions not based on a peer-to-peer model have been called the Product-Service Economy (Frenken, Meelen, Arets, & van de Glind, 2015). As stated by Europe Economics, 'while none of these terms are necessarily illegitimate, the sharing economy is by some margin the most commonly used and a reasonable descriptor which reflects common usage for the entire economic phenomenon we will study, rather than a part of it'.

1.1.1. Benefits and concerns of the sharing economy

Table nº 1: Benefits and concerns of the sharing economy

Benefits Concerns The enumeration and description of real and There is a lot of opposition to these sharing potential benefits of the sharing economy is not economy companies, especially coming from peaceful. Defendants will claim these and other traditional businesses and sometimes even benefits, while detractors will challenge them. We authorities that are threatened by the new therefore propose to read this list using the competition or by the new unregulated entrants to perspective of the advocates of the SE. the market. The legal framework of the sharing economy is unclear, and many are fighting to Affordable and qualitative products/services interpret existing laws in the context of P2P The entrance to the market of SE platforms has led sharing economy business models, and to the introduction of products and services at a considering whether new regulation is required. much lower price. People can now rent an entire flat at the price of a mid-range priced hotel room, Economic concerns return home by taxi for half the price of a normal By now, everyone agrees that the SE is shaking taxi fare, etc. grounds and disrupting the market. With their entirely new business models (basically there are Furthermore, a study by Wallsten (2015) on the facilitators rather than owners), SE platforms reactions by taxi companies in New York City and provide a real challenge to 'traditional' market Chicago reveals that the rise of Uber has caused players. These are losing market share and thus them to improve quality. Therefore, consumers revenue. Zervas & Proserpio (2015) have seem to be great benefiters. undertaken a study on the economic impact of Airbnb on the hotel market in Texas (USA) and Knowledge coordination and decreased found that in Austin, where Airbnb supply is transaction costs highest, the negative impact on hotel revenue lies The exchanges the SE facilitates are not between 8 to 10%. necessarily new (people have been renting out rooms, they have been sharing car rides, etc.) but While we acknowledge the relevance and the reduction in transaction costs because of importance of these concerns we will, following knowledge coordination allows for unlocking out mandate, focus on the legal concerns in the underutilized assets. paragraphs below.

PE 558.777 II - 132

Nowadays, many people have access to products Legal concerns and services they wish to purchase. However, we The key concern regarding the SE revolves around are facing a challenge that is related to data and the legal framework (or the lack thereof). This intelligence regarding what we want to buy. There section presents different subsets of this concern, is so much (often contradicting) information out which are taken from Susan Mclean's (2015) there that is has become cumbersome to make a article37. sound decision on what to purchase. Indeed, according to Allen & Berg (2014) it is not the cost Consumer Protection of the resource (asset) itself that constitutes a Operators need to consider the extent to which problem to society, but rather the cost of their platforms comply or not with applicable coordinating the knowledge that people need to consumer protections laws. While traditional transact the resource. People need to know how companies have to comply with the strict rules to exchange in order to speed up and increase that have been applied to them to ensure transactions. Such knowledge can lead to consumer protection, SE platforms have much decreased transaction costs. According to more freedom in this regard. A regular taxi Dahlman (1979), transaction costs consist of three company for instance, might be obliged to different types: undertake a background check on the criminal record of its drivers, while users of the application • search and information costs; Uber have long been exempted from this. • bargaining and decision costs; and Recently, the company has started to apply such • policing and enforcement costs. checks, however they are still contested38.

Making use of digital platforms, the SE is able to It is worth noting that some platforms add a 'plus' significantly reduce transactions costs. For of guarantees and security to their users, although instance, it has never been easier to gather not all of them do so. Another important element information on prices, quality, specifications, of to consider here is the value of the rating products and services. To be sure, we have fast systems/reputation scores, as an element of self- and cheap access to knowledge in the three protection by the users of the services. above-mentioned fields that compose transactions costs. SE platforms provide room for peer-reviews, Data protection rating systems, personalised searches, etc. This is a No different from many internet-driven pivotal factor for the success of the SE. companies, data protection has become an issue for some SE platforms as well. The platforms have the technological capacity to collect and store user Empowerment of economic actors data, and operators need to address issues of We have already touched upon this element in the compliance with applicable data protection laws, above paragraph but it still deserves some in terms of the processing of personal data of both explanation. Users of the SE play a key role in a users and users' customers, and prepare concept the SE has universally integrated: i.e. appropriate privacy policies and cookie notices. reputational mechanisms. Through reputational rating mechanisms, for example, it has become In case of transport and accommodation impossible for those that do not play by the rules platforms, it basically comes down to the to hide. Indeed, if users do not share in a following: they know where you are, how you got satisfactory manner, they will be excluded from there and where you will be. According to Rogers future exchanges. (2015), in the case of Uber, data privacy issues

37 Susan McLean, The rise of the sharing economy, Computers and Law magazine, Vol 26 Issue 1 April/May 2015. 38 http://www.cbsnews.com/news/the-rising-safety-issues-that-could-throttle-uber/ (last consulted on 25/10/2015).

PE 558.777 133

In addition to empowerment related to policing might self-correct since sale or exploitation of user and oversight, SE users are also empowered in the data is not considered a main source of revenue, sense that they receive additional earnings unlike for companies such as Facebook and alongside their regular income, and have the Google. If this is the case, then we might expect it opportunity to choose when and where they want to hold true for most of the SE platforms. to make use of platforms (which contrasts with traditional corporations where one has to follow Labour laws the rules regarding working time, work place, etc.). There is an ongoing discussion on weather SE Better (more sustainable) use of underutilized platforms should be considered information resources. providers or employers. The distinction has a major effect. If employer, a number of social rights This can be derived from the definition and it is are to be taken into account: Overtime, a arguably the pivotal benefit SE can provide to minimum salary, worker compensation society. Many of our assets remain underused. Yet (replacement wage and medical benefits during at the same time, we are facing huge medical leave), etc. One of the main issues environmental challenges. SE can help coming to a regarding the SE is then what employment status a more sustainable market. The French Environment user should have: employee or independent and Energy Management Agency (ADEME) has contractor? calculated that shareable goods account for about one third of household waste. Therefore, Discrimination intuitively, one can expect a positive impact from Operators need to consider potential the SE on the environment. Demailly & Novel discrimination issues. What would happen if users (2014) elaborate on the subject and explain how refuse to loan their car or hire their spare room three different elements could contribute to such based on the person's race, religion or sexuality? an impact. Would then the operator be liable under antidiscrimination laws? First, redistribution of goods is easy to understand to have a positive outcome to the environmental Laws relating to payments balance. If person X has a piece of furniture that If payments are made via the P2P platform rather he does no longer need and at the same time than directly between users, operators need to person Y is in need of such an item, redistributing address compliance with applicable payment rules the item (selling, exchanging or giving) has the and potentially deal with local payment services following results: laws. Clarity on who is responsible becomes very relevant. Nevertheless, many countries impose • Person X disposes his item and does not restrictions on certain types of payment structures necessarily need to replace it by another in order to protect the consumers. one • Person Y receives the item that he Taxation wanted without having to purchase a new Both operators and users might be subject to item taxation issues. For example, tourist taxes apply in • The lifespan of the item is extended, it many cities to hotel accommodations, and slows down the distribution of new therefore might be applicable to P2P models that products (that obviously need to be provide equivalent services. Collection of such produced, for which resources are taxes is an issue, where both operators and users required) of a given platform might be responsible. This is • In case person X had anticipated the obviously a major concern as it determines for a redistribution (that is to say, if he knew large part whether the SE is de facto a shadow that he was going to sell/exchange/give economy or not. away the item after a certain period of

PE 558.777 II - 134

time), this leads to the purchase and thus Safety and security the production of durable goods (since Security issues might be involved when operating one would only redistribute goods that via a platform. Some of them already request are still in acceptable condition)36 information on users to credit their positive records. In other cases, reciprocal reviews and a The second element is what they call system of ratings help to build trust among the 'mutualisation' and refers to short-term renting users. and lending of items. Here, the positive outcome derives from the following assumptions: Liability One of the key concerns is who is legally liable if • Renting or borrowing an item replaces something goes wrong: could the platform be held the need to purchase one responsible if a hired car crashes or a host's • The item that is rented out or lent out is apartment is damaged? of high quality and resistant to intensive usage Insurance • Repeated rental takes place at a local Some insurance companies are refusing to provide scale, minimizing environmental effects insurance if policyholders engage in P2P sharing. caused by motorized transport that At the same time, many users participate in the SE would have taken place if the item had without checking beforehand with their insurer been purchased. what is covered by the insurance and what is not. Industry specific laws The third element they discuss is shared mobility. While it is pointed out that cars are only used 8% Sharing economy platforms need to consider of their time and car sharing could therefore lead issues of compliance with any sector specific rules. to significant financial savings, the environmental The UK sharing economy sector for example is impact of car sharing can only be positive if: engaging in associations for their sector in order to • The shared car is more durable. This early engage with the regulators to design means that the lifespan of a car should common constructive solutions for the P2P not decrease as many times as the usage businesses. of the car increases due to sharing. • Users do not travel more distance with a shared car than they would have done with their own.

Provides workplace flexibility

According to Kumar (2015), SE responds to the growing demand for workplace flexibility. He argues that freelance work could be the employment of the future and SE has a vital role to play in it. For example in the case of drivers they may simultaneously use different platforms (Uber, Blablacar) to provide their services, without being committed to one single one, and combining it with their regular jobs.

Creates new services

PE 558.777 135

Still according to Kumar (2015), SE is responsible for the creation of many new services that could not easily be provided by traditional companies. As an example he refers to the case of Uber. A 'normal' taxi company could have never provided this type of service (basically a very cheap taxi ride for a great number of people) without having an enormous car fleet, which is extremely expensive.

Prevents the shadow economy Provided that mechanisms are in place to collect taxes from users of the SE that generate (part of) their income with their SE activities, the SE is a means to get people out of the shadow economy. It goes without saying that such an evolution has major benefits to the State Treasury and to society as a whole.

1.2. Research aims

While much attention is being given to the phenomenon of SE, its benefits and concerns, not much work has been produced to analyse the legal barriers which prevent the SE from reaching its full potential nor has there been a lot of analysis on what could be done about it at the European level. With this research paper we aim to close this gap in existing literature on the topic.

The paper aims to answer the following three research questions:

1. Are there legal obstacles/barriers that prevent the SE from reaching its full potential? 2. How effective is the existing EU regulatory framework in promoting and overseeing SE platforms? 3. What additional steps should be taken at EU level for the SE to realise its economic potential, with the objective of preserving a balance between creative freedom needed by business and the need for regulation of the sector? In this regard, are there any best practices from Member States that could be suggested? To what extent could self- regulation of the sector be a relevant option?

1.3. Methodology

To do so, the tasks to be performed are divided in two areas:

1. To take as a point of departure the draft report produced by Europe Economics, in order to avoid repeating what has already been done, and focusing on the identification of interesting issues that can further be studied and analysed in our report. We take as a point of departure the definition of the (SE) given by Europe Economics.

PE 558.777 II - 136

2. To analyse legal gaps for citizens, business and other relevant stakeholders regarding the sharing economy. To assess the effectiveness of the existing EU regulatory framework in this regard. To highlight best practices and provide recommendations as to possible EU action in this field.

Due to the novelty of the concept, we have chosen a phased funnel approach whereby we start with gathering as much information as possible and want to end up with a selected number of usable cases that allow for answering our research questions.

We therefore performed mainly desk research, using both primary and secondary sources. In this regard, in addition to the academic literature on the subject, special attention has been paid to cases brought before both national jurisdictions and the Court of Justice of the EU (hereafter, 'CJEU'). Furthermore, due to the novelty of the topic, a lot of research has been conducted on the Internet, using institutional and governmental websites, online newspapers and blogs. Finally, some interviews have been conducted with experts on the topic39.

Considering the time constraints we had we deliberately limited our analysis to six Member States (BE, DE, FR, ES, NL and UK). Even though these six proved to cover a wide array of national responses to the SE, it is quite possible that – especially because of the rapid rise of the SE, and together with it the concerns – other interesting cases have come up just before the publication. Furthermore, in order to respect the requirements, we choose to perform mainly desk research, using both primary and secondary sources. While a great deal can be found in books and on the Internet, it can be expected that nuances to national cases are more easily identifiable through other research methodologies, such as interviews for instance. Yet, these would have demanded a lot of time and such an approach would go beyond the objectives of this paper. It is nevertheless a suggestion for future research. On the same line, we have presented the court cases up a level we considered appropriate for the purpose of the study and to allow us draw to detect remaining legal gaps and draw conclusions. Again, it could also be interesting to analyse every case from every possible angle. However, this would leave us with hundreds of pages that only address specialists.

In light of the above, the report is divided in five chapters. In the first chapter, the introduction, the literature review on the SE, and the methodology are provided.

In the second chapter, the national legal responses given to the SE in BE, DE, FR, ES, NL and UK is addressed. This chapter covers the three sectors selected (transport, accommodation and professional services) analysing cases from different geographical areas. The chapter is divided per sector, and for each one of them, the following aspects are covered:

1. Study and compare the legal responses given by the Member States to the phenomena and the effectiveness so far of those responses. The legal responses analysed cover both the legislative and jurisdictional branches: in some MS the drivers of the responses have been triggered through Court cases, while in others it has been the legislator reacting to sector' protests without waiting for the problem to reach the Courts. 2. Comparative analysis of the different legal responses and identification of best practices.

In the third chapter, the EU regulatory framework is analysed. Questions relating to the Single market strategy, EU competence, and different areas of legislation of relevance for the purposes

39 Oui Share Festival, 19 November 2015, Barcelona: Interviews conducted with Albert Canyigueral and Miquel Ferrer, members of Oui Share Spain.

PE 558.777 137

of this report – in particular, the Services Directive, the eCommerce directive, consumer legislation, and data protection laws – are examined.

In the following chapter, the remaining legal gaps and issues, as well as the obstacles and barriers preventing the SE from reaching its full potential are analysed. We categorise through indicators the different obstacles and present a comparative analysis among the six member states analysed in the study. Finally we cover the potential of self-regulation in the SE.

In the concluding chapter, the need for EU action will be assessed. Based on the findings of the previous chapters, we will try to identify if and how possible EU action is needed. Recommendations are presented in this regard.

PE 558.777 II - 138

2. National legal framework: case studies

SE businesses are appearing at unprecedented rate; one can now find a sharing platform for almost anything. In order to define the focus of our research we have narrowed down the geographical area of investigation, the legal areas to be covered as well as the P2P sectors included in our analysis.

1. Geographical scope: we will concentrate the analysis of the cases geographically, based on a combination of factors like the geographical location (North /South); the amount of concentration of big cities where major platforms are strong; existing literature on the sharing economy in a given country, related to the amount of Court cases, relative size of the sharing economy users, social and citizens' mobilization around the phenomena etc. We are presenting in this briefing paper the analysis of cases relevant in BE, DE, ES, FR, NL, and UK. Where relevant, or in the absence of cases on the European continent, some examples were taken from the United States of America. 2. The law areas covered: we will highlight the legal issues that are raised in the cases presented in this report. With special attention to the questions related to Labour law, Consumer protection (including insurance) and Fiscal law. 3. The platforms to be analysed. Because of the discussions and debates they are causing in the media and academic and political spheres we have focused our analysis on platforms that operate in the following sectors: (1) Transportation; (2) Accommodation; (3) Professional Services. Due to the time constraints and limited budget, the final number of platforms analysed in the research is limited.

2.1. Transportation

Uber has given rise to the most notable example of legal conflict arising out of SE platforms in the transportation field. Blablacar is also a SE platform for transport with a different business model that has been also subject in some MS to different legal disputes; we will therefore focus our analysis upon the legal issues surrounding Uber (and when relevant also Blablacar) in six Member States.

Uber This business model allows consumers with smartphones to submit a trip request which is then routed to Uber drivers who use their own cars. Uber fixes the price, and occasionally they use a dynamic pricing model at peak seasons/moments, to encourage drivers to be available: With some events (New Year 's Eve, a storm), the car demand increases while the car supply tends to shrink. In these cases the supply curve is moving left at the exact same time that the demand curve is moving right. As a result Uber considers vital to use price as a catalyst to increase supply.

By May 2015 the transportation network company, founded in 2009 and headquartered in San Francisco, was giving service in 58 countries and 300 cities worldwide.

PE 558.777 139

According to Uber owners, Uber's drivers are independent agents that are either self-employed, or work for someone who owns multiple cars. Uber does not own cars and they claim not to employ drivers: the drivers decide whether or not to open the Uber application and accept requests for rides from Uber customers. The drivers are not bound by exclusivity, and many of them work on multiple services, and many have 'regular customers' that they engage off the Uber platform.

The majority of Uber fares go to these independent drivers: on average, over 80% of gross fares. Uber Company argues that of the percentage that is retained by Uber, a large portion goes to cover variable expenses within the service. These expenses include payment processing, payment fraud, refunds, customer support, dispute resolution, cellular handsets and service fees for the drivers, and local regulatory efforts. They claim to have a low margin business40.

As Uber grew internationally, it began to experience disputes with governments and taxi companies. In Europe it has been banned in different cities, with different legal arguments.

The company has requested an investigation to the European Commission to see whether some of these restrictions are legal and do not violate EU law. At least Spain, France and Germany are on the list according to different information.41

Blablacar The platform connects drivers and passengers willing to travel together between cities and share the cost of the journey. Prices are fixed when a car owner offers a ride42. They are non- negotiable and the same for all co-travellers. The price is always based on a suggestion calculated by BlaBlaCar according to the itinerary and real costs incurred by car owners. Car owners are free to adjust the price within reason, to account for the comfort of their car or their willingness to make a detour. The price cannot exceed a ceiling set by BlaBlaCar, in order to ensure costs are fairly shared and that car owners do not make a profit. Generally, price per seat represents a third of the fuel cost for the journey; when a car owner takes three co-travellers, he offsets all fuel costs, but not necessarily tolls. Payment is made through the application: the passenger pays online at the moment of the reservation, and the driver receives the payment once the journey is completed. The platform charges since recently a percentage of the price.

BlaBlaCar has more than 20 million members across 20 countries. Before the on-line payment, BlaBlacar (ex-covoiturage.fr) has made his living with web ads and selling custom ride sharing platforms for companies (e.g. Ikea, Carrefour) and cities. But the B2B model was not profitable, requiring too many customisations for lower and lower costs, so the founders focused entirely on the C2C model and looked on how to improve it and monetise it. Since mid-2012, they deployed a booking system and then charge a few fees when people book a ride through their

40 https://newsroom.uber.com/guest-post-a-deeper-look-at-ubers-dynamic-pricing-model/ (last consulted on 25/10/2015). 41 http://techcrunch.com/2015/07/15/following-uber-plea-european-commission-investigates- germanys-restrictions/ (last consulted on 25/10/2015). And http://tech.eu/brief/judge-uber-european-court-of-justice/ (last consulted on 25/10/2015). And www.reuters.com/article/2015/04/01/us-uber-eu-complaint-idUSKBN0MS4BH20150401 (last consulted on 25/10/2015). 42 https://www.blablacar.in/faq/question/how-do-pricing-and-payment-work (last consulted on 10/01/2016).

PE 558.777 II - 140

site. Since the introduction of such system, BlaBlaCar's revenue has been increasing each year and the company has developed to become a European market leader in ride sharing (current leader is Carpooling in Europe).

2.1.1. Benefits

Unlike employees of taxi companies, drivers of Uber chose when to drive and hence to work. This provides drivers with the opportunity to earn an income without the strict full-time employment scheme. Furthermore, there is more efficient use of resources, in this case the car and time. While traditional taxi drivers often have to wait around at taxi stands for the next passenger, Uber drivers can use their time as they like, possibly filling their spare time offering themselves out for different sharing services.

Consumers also benefit from Uber in several ways. First, Uber rides complement public transport, which is often limited in time (no 24/7 service) and space (no door-to-door transport). Second, they can offer a cheap alternative to traditional taxi rides.43 Finally, the convenience of getting an Uber ride adds value. A recent study comparing Uber with traditional taxi services in the Los Angeles area44 reveals that UberX rides arrive in less than half the time of telephone-dispatched taxi.

2.1.2. Complaints

The rise of Uber has been the instigator for protest in many big cities all over the world. These protests are led by traditional taxi companies and drivers. They argue Uber is challenging them with unfair practices. More concretely, Uber is offering the same type of services without being subject to the same rules, the argument goes. These rules deal with among others, social benefits, working time, taxes and insurances.

2.1.3. Case analysis

Belgium

In Belgium, it was not a taxi company that took Uber to court but one that in a way offers the same type of service as Uber. The Brussels-based company 'Taxis Verts' (Green Taxis) assumes the role of contact centre between customers and affiliated taxis. Taxis Verts itself is therefore not subject to the rules regarding taxi services as laid down in a Decree of 1995.45 However, professional taxi drivers to whom it assigns taxi rides do have to comply with the rules. Taxis Verts claims Uber is offering exactly the same services as themselves since it offers taxi services against payment, at a moment where Taxis Verts itself is offering a web application to its customers to request a service. Summarised, Taxis Verts blames Uber to commercialise a service

43 http://www.telegraph.co.uk/technology/uber/11902613/Proof-that-Uber-costs-less-than-black- cabs.html (last consulted on 25/10/2015); http://www.csmonitor.com/Business/Saving- Money/2015/0626/Is-Uber-cheaper-than-a-taxi-Not-always (last consulted on 25/10/2015). 44 Smart et al (2015). 45 Ordonnantie betreffende de taxidiensten en de diensten voor het verhuren van voertuigen met chauffeur.

PE 558.777 141

in the framework of rides that fulfil the criteria of a taxi ride, delivered by drivers that do not possess the required license nor comply with the rules. It therefore accused Uber of unfair competition with regard to drivers with the necessary documents as well as to other companies such as Taxis Verts, to which they are affiliated.

Uber, in a response to this action, claimed their users indeed receive a payment, but this should be seen as a compensation for the costs made rather than a form of salary. The Brussels Tribunal of Commerce (Rechtbank van Koophandel) did not follow Uber's explanation and ruled that Uber is indeed offering unlicensed taxi services against what can be seen as a salary, since the payment offered to drivers can exceed the costs made by the ride. It therefore imposed a ban on the service UberPOP in Brussels.46 Nevertheless, the Tribunal decided to refer a question for a Preliminary Ruling on whether this strict interpretation of the Decree would not interfere with articles in the Charter of Fundamental Rights of the European Union, the Treaty on the Functioning of the EU and the European Charter of European Rights. In essence, it boils down to the following question: Does the principle of proportionality as laid down in the above mentioned documents interfere with the rules as written in the Brussels Decree on taxi services for those cases where 'taxi services' would be applicable even to occasional, not compensated for in the form of a salary, rides as a reaction to requests for taxi services?

In the meantime, the Government of the Brussels Region is developing a legal framework for alternative taxi services, such as Uber. In consultation with the taxi sector it is creating a framework for all types of paid transport, in order to abolish unfair competition and social dumping. The former Brussels Minister of Transport, Pascal Smet, has said that this does not constitute a legalisation of Uber. However, if companies like Uber abide by the rules concerning transparency, safety, liability as well as the social and fiscal rules, they can get a legal base for their operations. Notwithstanding this legal base, Uber would not enjoy the same privileges as regular taxi companies. Drivers could not work full time, they could not make use of priority lanes nor could they pick up clients on the streets. It is the Minister's expectation that Uber will have to increase its prices if it wants to follow the rules (De Redactie, 2015).

Germany

In Germany, Uber is facing problems with several local authorities. It started on 25 August 2014, when UberPOP was prohibited by a preliminary injunction of a district court in Frankfurt am Main (Hessen).47 Drivers without an official permission, according to the Personenbeförderungsgesetz (1964), were not allowed to offer their services. The company could be fined with a maximum of € 250.000. The reason for the lawsuit was a complaint from the German organisation for taxi companies, Taxi Deutschland Servicegesellschaft für Taxizentralen.

Three weeks later, the same court lifted the preliminary injunction. The German taxi organisation that asked for the court decision had based its demand on reasons of urgency –

46 www.ie-forum.be/www.delex-backoffice.nl/uploads/file/IEFBE/IE- Forum_nl%20Voorz_%20NL%20Re chtbank%20van%20Koophandel%20Brussel%2023%20september%202015,%20IEFbe%201541%20(Uber %20BVBA%20tegen%20Taxi%20Radio%20Bruxellois).pdf (last consulted on 10/01/2016). 47 Landgericht Frankfurt am Main Beschl. v. 25.08.2014, Az.: 2-03 O 329/14.

PE 558.777 II - 142

which in itself was unjustified because Uber already existed since 2013. The content of the injunction itself was not contradicted, but the procedure should have been different.

Uber's fortune did not last long. On 26 September 2014 the local courts of Hamburg and Berlin ratified the first ruling of 'Frankfurt'. In their jurisdiction UberPOP was prohibited.48 The motivation was based on the fact that the consumer rights and safety were not sufficiently guaranteed. In the case, Uber was not merely an intermediary party; it represented the drivers. Uber appealed, without success. On April 16, 2015 the initial court decision was upheld by the Berlin-Brandenburg Court of Appeal.

In March 2015, the Frankfurt Regional Court issued a nationwide ban against Uber ride-hailing service UberPOP (UberBlack and UberTaxi were not affected), declaring its business model illegal. Mediating rides with private drivers who don't have the required licenses is illegal.49 Uber tried to convince the court that the price the customer had to pay was no more than the transportation costs. The court, however, argued that the price per kilometre charged by Uber, was far higher than the actual costs. The surplus goes to Uber (20%), Dutch taxes (Uber operates from The Netherlands), and an income for the drivers. Thus, the court argued, Uber has to be seen as a regular business, without the necessary licences. Also in this case, Uber appealed. Uber decided to continue operations in several cities, but only with its limousine service (UberBlack) and licensed taxi drivers (UberTaxi) with lower fares than regular taxis. However, due to a lack of drivers, in November 2015 Uber suspended services in three major German cities: Hamburg, Frankfurt and Dusseldorf, and retreated to Münich and Berlin. 'For many prospective Uber partners the process of registering an independent rental car enterprise has proved as too costly and time consuming,' Uber said in a statement. It added it would improve its services in the two remaining German cities and 'intensify the dialogue' with lawmakers and authorities, saying Germany remained one of its most important global markets.50

At the moment of writing this report, UberPOP is still banned in Germany. Uber needs to hold a taxi operator's license and comply with all the existing taxi legislation. An appeal case regarding the court ruling of the Frankfurt Regional Court is still pending.

France

In the first half of 2014, the UberPOP version of the app was launched in Paris, France, whereby users are linked to drivers without professional taxi or chauffeur licenses, while Uber covers supplemental insurance.

While the public response to the SE has been quite positive in France, the creation of web applications such as UberPOP has given rise to important social problems and mobilizations from the traditional actors in the transport sector, as they consider that SE platforms exploit the existing legal gaps to infringe upon the principle of fair competition. The stance adopted by the French authorities in this field has been quite conservative.

48 Berlin, see: http://www.berlin.de/sen/justiz/gerichte/vg/presse/archiv/20140926.1255.399438.html (last consulted on 10/01/2016). 49 Landgericht Frankfurt am Main Urt. v. 18.03.2015, Az.: 3-08 O 136/14. 50 http://www.reuters.com/article/us-uber-germany-cities-idUSKCN0SO2R620151030 (last consulted on 10/01/2016).

PE 558.777 143

As a preliminary comment, it should be noted that the transportation market is divided into two domains: the cruising market, which is reserved to taxis, and the advanced booking market, which is open to competition and therefore includes chauffeur-driven vehicles. The legal monopoly of taxis in the cruising market is justified on public interest grounds, relating to the policing of traffic and parking. In return, taxis are subject to regulated prices and to a system of administrative licence authorizing them to stand at rank within their municipality or the common service to which they are attached. By way of example, this licence amounts to 230.000 Euros in Paris51. Nonetheless, technological innovations and the appearance of SE platforms in the transports industry such as Uber blurred this distinction, and allowed the fast development of chauffeur-driven vehicles52.

Prohibition on electronic cruising; obligation to return to base; pricing system

Against this background, the French legislator has intervened to regulate this activity, mainly through the adoption of the so-called 'Loi Thévenoud'53. This Act recalls the prohibition of cruising in a street open to public circulation to chauffeur-driven vehicles, thus seeking to guarantee the legal monopoly enjoyed by taxi drivers. In particular, it limits the use of GPS by chauffeur-driven vehicles, and it prohibits electronic cruising through smartphones, as well as the solicitation of clients without advanced booking, except for taxi drivers54. The prohibition on cruising is punished by criminal sanctions.

In addition, it regulated chauffeur-driven vehicles pricing system55: The total price of the service needs to be determined at the time of the advanced booking. Therefore it was prohibited to charge on the basis of both time and distance. This was considered by the plaintiffs to violate the freedom of enterprise.

Finally, it established an obligation to return to the base applicable to chauffeur-driven vehicles: drivers must return to the establishment of the operator or to an off-road location where parking is authorised, unless (s) he is able to provide evidence of a prior reservation or a contract with the client56.

Uber France SAS and Uber BV challenged these legal barriers before French courts57. They argued that these articles infringed upon the freedom of enterprise, the principle of equality

51 Conseil constitutionnel, Commentaire Décision nº 2015-468/469/472 QPC, Société Uber France SAS et autre, p. 3. 52 Conseil constitutionnel, Commentaire Décision nº 2015-468/469/472 QPC, Société Uber France SAS et autre, p. 3. 53 Loi n° 2014-1104 du 1er octobre 2014 relative aux taxis et aux voitures de transport avec chauffeur, « Loi Thévenoud », JORF n°0228 of 2 October 2014, p. 15938. Available in French at: www.legifrance.gouv.fr/eli/loi/2014/10/1/2014-1104/jo/texte (last consulted on 26/10/2015). 54 Transports Code, article L. 3120.2 (III). 55 Transports Code, Article L. 3122-2. 56 Transports Code, Article L. 3122-9. 57 French Court of cassation, commercial chamber, judgment no. 376 of 13 March 2015. Available in French at: www.legifrance.gouv.fr/affichJuriJudi.do?oldAction=rechJuriJudi&idTexte=JURITEXT000030356182& fastReqId=1459466973&fastPos=2 (last consulted on 25.10.2015); French Court of cassation, commercial chamber, judgment no. 375 of 13 March 2015. Available in French at:

PE 558.777 II - 144

before the law and the right to ownership. The French Constitutional Court (Conseil constitutionnel) held that both the prohibition on electronic cruising and the obligation to return to the base are constitutional while the regulation of the pricing system of chauffeur-driven vehicles is unconstitutional58.

Prohibition of UberPOP

The Loi Thévenoud introduces a distinction in the Transports Code between Uber-like platforms and Blablacar-like platforms; the former are considered as a disguised form of car sharing and are illegal whereas the latter is legal, as they are understood as real forms of SE. An amendment to the Transports Code was made in 2015 in order to clarify this difference. Car sharing is defined as the common use of a motorized ground vehicle by a driver and one or more passengers, provided that the costs are merely shared among them and the trip is carried out by the driver for their own purpose59. In contrast, platforms like UberPOP, that connect clients with people who engage in individual transportation activities without being transport companies, taxis or chauffeur-driven vehicles, are illegal and subject to criminal sanctions60. The Constitutional Court held a judgment on 22 September 2015 whereby it concluded to the constitutionality of UberPOP's ban61. It is worth noticing that the Constitutional Court confirmed the difference of treatment between Uber and Blablacar, as it considered that the challenged provision did not aim to prohibit car-sharing platforms62.

Two CEOs of Uber were arrested on charges of being illegally handling data and illegally operating a taxi service; they will be judged together with Uber France in 201663.

The Netherlands

The Netherlands has three sorts of the Uber app. The drivers Uber Black and UberLUX are certified drivers that use Uber in addition to their other duties and services. Legal but not undisputed. With the third one, UberPOP, the drivers are individuals. The controversial

www.courdecassation.fr/jurisprudence_2/qpc_3396/375_13_31319.html (last consulted on 25.10.2015); French Constitutional Court, Decision 2015-468/469/472 QPC, 22 May 2015, Société UBER France SAS et autres. 58 French Constitutional Court, Decision 2015-468/469/472 QPC, 22 May 2015, Société UBER France SAS et autres. 59 Transports Code, article L. 3132-1. 60 Transports Code, Article L. 3124-13. (« Est puni de deux ans d'emprisonnement et de 300 000 € d'amende le fait d'organiser un système de mise en relation de clients avec des personnes qui se livrent aux activités mentionnées à l'article L. 3120-1 sans être ni des entreprises de transport routier pouvant effectuer les services occasionnels mentionnés au chapitre II du titre Ier du présent livre, ni des taxis, des véhicules motorisés à deux ou trois roues ou des voitures de transport avec chauffeur au sens du présent titre » ) 61 French Constitutional Court, Decision n° 2015-484 QPC of 22 September 2015, Société UBER France SAS et autre (II) Available in French at: www.conseil- constitutionnel.fr/decision/2015/2015484qpc.htm (last consulted on 25/10/2015). 62 Benjamin de Sevin, 'Uberpop et BlaBlaCar', La Semaine juridique. Administrations et collectivités territoriales, 5 octobre 2015 (n° 40), p. 1-2. 63 www.lemonde.fr/entreprises/article/2015/09/30/deux-dirigeants-d-uber-france-devant-la- justice_4777645_1656994.html (last consulted on 25/10/2015).

PE 558.777 145

question has been – like in other countries-cities, to what laws and regulations the individuals offering their services must comply.

Uber contributed to the discussion by the start of a pilot with UberPOP in Amsterdam, and gave order to the consultancy firm Accenture to map how many people use the service, what their motives are for this and how they experience the service in terms of flexibility, safety and cost. In October 2014 the evaluation of Accenture was presented. Although the test officially had ended, Uber continued to function. Meanwhile the Dutch Inspection Environment and Transport controlled regular taxis on the compliance with the Law on Personal Transport 2000 (WP2000), and fined them if anything was not in order.

The report showed that most of the drivers had other jobs, and saw Uber activities as an extra income. One of the basic conclusions was that Uber deals with ordered personal transport and no 'hitchhiking' service like Toogethr and Blablacar. Respondents did not mind that Uber- drivers were not official taxi drivers. Official taxi drivers in Amsterdam have a poor reputation. Basically, Uber-users were quite satisfied with the service.

The report was sent to several MPs, in the hope to influence discussions on new taxi laws. According to the Dutch Inspection Environment and Transport anyone who offers taxi transport needs to comply with the WP2000 and have a licence. Uber says it does not offer transport, it only offers an application that enables others (the drivers) to offer their services. And therefore it does not need a licence.

The legislator did not agree: offering taxi transport without a licence in NL will be punished with a fine of € 4300. Also the car can be taken into custody. If the same driver is caught again, he will be fined with a € 10.000 (a maximum of € 40.000). As well a fine will be given to Uber (up to a maximum of € 100.000). Several Uber drivers were caught and fined.

In order to stop this procedure, Uber went to the Dutch Commercial Court of Appeal, asking for a preliminary injunction.

In December 2014 (Judgement: ECLI:NL:CBB:2014:450 (8 December 2014)) the court denied the request64, and argued: ' The Minister is right when it argues that the drivers have violated the prohibition of the provision and supply of unlicensed taxi. And that Uber is the co-perpetrator of this violation by the drivers'. The judgment can be interpreted this way that Uber is in fact a provider of taxi services. And not just simply as a broker or other intermediary that should perhaps be excluded from the authorization. The judge ruled that between the drivers and Uber there is 'conscious and close cooperation' to provide taxi services. Uber selects the drivers, the drivers can only access the app through Uber, Uber determines the fares and Uber receives a percentage of the fare. In short, Uber will not get away with the claim that it only offers an app that brings together supply and demand of individuals.

The question that has been arising in the social media is: why would an individual who by his own car brings another private person from A to B, where supply and demand is established via an app, must have a taxi license? What interest does this serve? According to the Minister the regulations on taxi licenses is aimed at 'a minimum required transport quality' and the

64 Available in Dutch at: http://uitspraken.rechtspraak.nl/inziendocument?id=ECLI:NL:CBB:2014:450 (last consulted on 25/10/2015).

PE 558.777 II - 146

interests of consumers. But does this actually lead to a form of service that should be controlled by government regulations? Or is it more a question of the convergence of supply and demand of clients, offering private individuals arrange their affairs, albeit with help from a commercial third party? A sort Marketplace? Does the government have a role in this at all?65 On the other hand, - social media debates argue - we need to consider the problems of traffic congestion in big cities, as well as environmental considerations to favour public instead of private transport.

In this context, Minister Henk Kamp promised 'technology-neutral regulations' that would ensure no company is disadvantaged because it is using different technologies. This would, for example, allow taxi drivers to replace their meters with smartphone apps like Uber's.

A government spokesperson said UberPOP would remain illegal. 'We want to make the taxi market more open, but we need to have fair competition,' spokeswoman Karin van Rooijen said.

Despite the judgement of December 2014, and even though Uber started a so-called 'bottom- procedure', in Amsterdam and Utrecht Uber drivers have been caught without having any licence. Uber was getting 'underground' by offering its services only to known customers, not accepting any new customers. In March and September 2015, police officers entered Uber offices in Amsterdam to get hold of Uber administration.

Finally on 20 November 2015, Uber decided to stop with UberPOP in the Netherlands66.

Spain

At the end of 2013, Uber started operating in Spain its UberPOP taxi service, which caused taxi strikes in Madrid and Barcelona 16 months later. After a ban from Berlin, Spain joins the ban and in December 2014 Uber suspends its taxi service in Spain, to set up three months later UBER EATS in Barcelona67. In Madrid and Barcelona parallel cases were brought to the tribunals on unfair competition grounds. Both cases have followed different paths:

In Madrid, the association of taxi drivers decided to start proceedings against the company. The resulting ruling establishing precautionary measures argued that the drivers contracted by Uber do not have the required administrative license to provide the service. Furthermore, the activity constitutes – according to the precautionary measures – unfair competition, (the company had not been certified as transport mediator in Spain) ordering to the telecom companies and electronic payments, to ban any transaction and hosting of the company UberPOP. At the moment of writing this article, the precautionary measures have been prolonged though with

65 http://njb.nl/blog/uberpop-(2).13294.lynkx (last consulted on 25/10/2015). 66 'Today we have informed all Dutch drivers that on Friday November 20 at 12:00 am we stop UberPOP in the Netherlands. Despite the fact that UberPOP has led to interesting insights about how the mobility of the future might look like, we decided to stop UberPOP. We believe that continuing is an obstacle to a constructive dialogue on modernization of existing laws and regulations. The services UberX, Uber Black and UberLUX are not in question and remain available.' See: https://newsroom.uber.com/amsterdam/nl/uberpop-stopt-in-nederland/ (last consulted on 13/01/2015). 67 For information on UBER EATS Barcelona, See: 'Despite Court ban in Spain, Uber drivers return to streets in Barcelona', El País, 15 July 2015.

PE 558.777 147

some modifications: the unfair competition affects only UberPOP, making possible the access to other non-related companies.

In parallel, a claim was brought before the same Judge, by the Confederation of Buses (Confebus) against Blablacar Spain68. The plaintiff is using similar arguments to the ones of the Taxi confederation against Uber, and has requested provisional measures to have the web page closed. At the moment of writing this article the case is waiting for a verdict.

In Barcelona, the Judge referred the case to the CJEU asking whether Uber should be considered a transport company or an information society service69. Although the ruling will not be solved before the summer of 2016, a lot of expectation has been raised already: if the Court decides that Uber is a digital service provider, it will be much harder for national regulators to curb its activities. But if it were considered to be a transport service, it would increase significantly its operational costs. This could include how drivers are insured and its approach to adhering to local correspondent70.

The specific questions referred to the CJEU are the following:

1. Whether Uber is a 'mere transport activity', and is therefore excluded from the scope of Article 2.2.b of Directive 2006/123/EC of 12 December 2006 on services in the internal market; 2. If not, whether Spanish unfair competition law as applied to 'information society' services is contrary to European law, specifically Article 9 of the Services Directive, which states that an authorization, licensing or permits regime cannot be restrictive or disproportionate, and cannot unreasonably hinder the principle of freedom of establishment.

All legal proceedings against Uber in Spain have now been suspended, pending the ruling of the CJEU. For a deeper analysis of the questions posed in the preliminary ruling see Annex II.

The United Kingdom

In September 2015, Transport for London, the government body responsible for the transport system in Greater London has put forward proposals for private hire companies, such as Uber. Consultations on these proposals were held until 23 December 2015. According to our knowledge, no final decision has yet been taken. A number of these proposals could hit Uber's operations in London. In particular, the proposal to install a five-minute wait time between the ordering of a taxi and its arrival would have a huge impact since many of Uber trips took place within five minutes of the car being booked.71 Other proposed rules include a ban on showing

68 www.blablacar.es/. 69 www.poderjudicial.es/cgpj/es/Poder-Judicial/Tribunales-Superiores-de-Justicia/TSJ- Cataluna/Noticias-Judiciales-TSJ-Cataluna/El-Juzgado-Mercantil-3-de-Barcelona-plantea-cuestion- prejudicial-al-Tribunal-de-la-UE-en-relacion-al-conflicto-de-UBER (last consulted on 30/10/2015). 70 Murad Ahmed, European Technology Correspondent, Financial Times, 20 July 2015. 71 http://www.cityam.com/227908/transport-for-londons-new-five-minute-rule-for-uber-and-other- taxi-a pps-will-cost-drivers-19m-a-year (last consulted on 30/10/2015).

PE 558.777 II - 148

cars for hire within a smartphone app, one of Uber's main features, and the ability to book rides up to seven days in advance, something Uber does not offer. Furthermore, drivers would only be allowed to work for one operator at a time, while Uber states a great deal of its users work for traditional minicab firms too. Finally, the draft proposal suggests there should be controls on ridesharing in public vehicles. This would potentially harm the service UberPool, launched only in December 2015 in London that allows several customers to share a car.72

At the same time, Transport for London asked for judicial review on Uber's operations. The key question was whether the smartphone app should be considered a taximeter, which is prohibited in private hire vehicles.73 The argument held by the Licensed Taxi Drivers' Association was that the way Uber calculates fares on smartphones is technically metering. However, the judge ruled in favour of the taxi-hailing company. In the judgement, it is written that:

'A taximeter, for the purposes of Section 11 of the Private Hire Vehicles (London) Act 1998, is not a device which receives GPS signals in the course of a journey, and forwards GPS data to a server located outside of the vehicle, which calculates a fare that is partially or wholly determined by reference to distance travelled and time taken, and sends the fare information back to the device.'74

Therefore, the smartphone with the Uber app does not constitute a breach of the taximeter prohibition. In addition, the judge declared that it was drivers, not cars, who are equipped with smartphones.75 While Uber is seen to have won this battle, it remains to be seen what the new regulations on private hire companies will look like and whether the Licensed Taxi Drivers' Association will appeal the Court's ruling on Uber's taxi meter.76

USA

A case in California is worth mentioning, as it highlights very well the labour implications of applications like Uber. In this regard, the California Labor Commission ruled in favour of an Uber driver who alleged that she is an employee, not an independent contractor77. Uber decided to appeal the decision, which is still pending. Even though this case takes place outside of Europe, it arguably highlights a trend, that should seriously be taken into account by European authorities.

72 http://www.theguardian.com/business/2015/nov/30/uberpool-service-londoners-share-taxi-stranger (last consulted on 30/10/2015). 73 http://www.bloomberg.com/news/articles/2015-10-02/london-transport-asks-court-to-rule-on-uber- s-use-of-phones (last consulted on 30/10/2015). 74 https://tfl.gov.uk/info-for/media/press-releases/2015/october/high-court-declares-that-smartphones- used-in-private-hire-vehicles-are-not-taximeters (last consulted on 30/10/2015). 75 Ibid. 76 http://www.bbc.com/news/uk-england-london-34549700 (last consulted on 30/10/2015). 77 Uber v. Berwick, California Labor Commission, June 2015. Available at: http://es.scribd.com/doc/2 68980201/Uber-v-Berwick-California-Labor-Commission-Ruling (last consulted on 12/01/2015).

PE 558.777 149

2.2. Accommodation

Already since some years people have rented out their apartments through their own websites, but also through platforms, new companies offering space in return for a percentage of the letting price. The biggest ones like HomeAway, Housetrip (both US-based) have some million apartments and houses for rent. Of course their growing rate is limited, because only a small percentage of the internet-users have a spare apartment to rent out.

Airbnb (USA) has raised this business to a new level. Main offerings are not only the whole house/apartment, but also bedrooms. This makes 'sharing' the house, an option for almost anyone. Latest statistics show a rapid increase in offerings. Touristic cities like Berlin, London, Amsterdam and Barcelona have all more than 10.000 rooms and apartments to offer through Airbnb while Paris reaches a staggering number of 40.000.

Airbnb is not the only company in the market. Wimdu (Germany) is showing similar growing rates. Many smaller websites with similar offerings are popping up.

Table nº 2: Number of apartments available per city 2015

State of Company Paris London Berlin Barcelona Amsterdam Brussels World origin Airbnb USA 40.000 20-30.000 12.000 12.000 11.000 - 2.000.000 HomeAway USA 8.000 4.500 1.900 5.000 1.700 - 1.000.000 Housetrip UK 4.500 - - - - - Wimdu Germany ------300.000 Flipkey USA ------300.000

Source: Own elaboration

2.2.1. Benefits

The benefits pointed out by Airbnb seem to be obvious too.

For the platform it benefits from unique marginal economics: while a hotel chain expands by creating more hotels it has high marginal costs of maintenance. For Airbnb expansion means almost zero costs: new rooms are added and maintained by hosts. Airbnb does invest in community management to ensure that hosts follow best practices, and it offers insurance cover to hosts to encourage them to participate in the platform. But compared to a hotel chain, the marginal costs of creating added value are much lower.

For small house owners (the providers), the platform offers the possibility of renting out space to a huge market, which allows them to pay domestic household bills, an argument worth a million in times of economic crisis.

PE 558.777 II - 150

For tourists (the users) the platform offers competitive process compared to hotel rooms, and opens up a new way of interacting with the locals by opening up the possibility to use the local houses.

For cities, the benefit comes from the spending of tourists who otherwise would not have visited the cities involved.

Following the trend in many US cities, in Europe legal disputes are growing. As a response public authorities are approving new rules – like in the case of Berlin – to create new housing laws banning regular short-term renting of rooms without permission from the authorities, or otherwise supporting the trend but under clear regulatory conditions.

2.2.2. Complaints

Like in the case of Uber, the holiday rental site Airbnb finds itself under growing attack from city authorities. Airbnb provides an online platform to allow individuals to rent out their homes, rooms or apartments to visitors. While this is not in itself illegal, in most cases it breaches local housing laws, as well as fiscal obligations stemming from the activity.

As a consequence, complaints are being heard from traditional economy sectors. 'Unfair competition' is the most common. Hotels, Bed Breakfasts, and other lodgings have to register, pay taxes and comply with a lot of regulations. The owners who rent private rooms via companies like Airbnb usually are not registered, do not pay taxes, and do not comply with any regulations.

Another complaint – coming from citizen's platforms – deals with the supposed withdrawal of rental space in a city, and the forthcoming rise in rents. Since the business of renting out tourist apartments and rooms is getting more and more profitable, part of the offerings of Airbnb come from companies that manage several (sometimes even tens) of apartments at the same time, apartments used for tourists, and not for house-seeking locals.

These complaints – even though it sounds logical – are not supported by statistical evidence. Research in the case of Berlin (www.Airbnbvsberlin.com) did not come with solid conclusions. A final complaint is related to the fact that cheap rentals attract too much rowdy tourism.

2.2.3. Case analysis

Belgium

Brussels The Brussels Regional Government will impose a new rule as from 2016 for touristic accommodations. The scope of the regulatory framework will be extended from hotels and guest rooms to rooms offered by residents (which constitutes the largest part of Airbnb users). From 2016 onwards, Airbnb users offering a room will need a registration number. In order to receive one, they need to provide the following documents: a copy of proof of ownership, an excerpt from the criminal record, an insurance contract, a certificate of fire safety, and a plan of

PE 558.777 151

the property (De Morgen, 03 November 2015; Haeck, 2015). While a normal procedure for getting a certificate of fire safety requires an inspection from the fire department, individuals will only need to present an inspection certificate of gas and electricity (Haeck, 2015). All of these documents will have only to be presented once (Haeck, 2015).

Flanders The Flemish government is reviewing78 the Lodging Decree79 by which it wants to offer a framework to govern sharing economy platforms such as Airbnb. Rules will be simplified and clarified. A license or registration would no longer be necessary to offer a room for rent as long as all safety provisions are complied with. Whoever wants to be recognized as an official touristic accommodation operator can voluntarily request a recognition, in which case an inspector will pass by (Cludts, 2015). Online platforms such as Airbnb are also considered in the Decree and will have to provide the Government with addresses so targeted inspections can be performed (Goddefroy, 2015).

Wallonia The Walloon Government has not yet taken any measures in response to Airbnb's growth. It has to be noted that the platform has only got a small number of users in the French speaking part of Belgium (Godart, 2015). The current legislation requires from people offering touristic accommodation that they identify themselves in front of the municipal administration where the accommodation is situated and that they possess a certificate of simplified control (for dwellings offering space to less than 9 persons) or a fire safety certificate. This is the only requirement to enter the touristic housing market (Le Vif, 13 July 2015).80 One could argue here, of course, about the definition of touristic accommodation and question whether Airbnb users voluntarily identify themselves or whether the government would take action to detect unidentified users.

The Brussels and Flanders cases constitute two different approaches taken by the different regions in Belgium. While Brussels choses to impose more rules, obligations and controls on residents willing to offer a room for rent, Flanders opts for an opposite, more liberal, approach with controls to be performed a posteriori. The Walloon government's non-approach, then, could be seen as leaning towards the Flemish approach, since it requires only one certificate.

Germany

Airbnb: The number of Airbnb listings in Germany is changing every month. What is clear, is that Berlin has by far the biggest offering: somewhere between 10.000 and 15.000 apartments. Complaints of local citizens and politicians about the impact on local rental prices, led to an investigation report from students from the University of Applied Sciences in Potsdam,

78 http://www.toerismevlaanderen.be/logiesdecreet (last consulted on 30/10/2015). 79 Departement internationaal Vlaanderen Dienst Toeristische Vergunningen (2014) Het Vlaamse Toeristische Logiesdecreet en uitvoeringsbesluiten, available at: http://www.vlaanderen.be/int/sites/iv.devlh.vlaanderen.be.int/files/documenten/Geco%C3%B6rdi neerde%20versie%20van%20het%20Toeristische%20Logiesdecreet%20en%20de%20uitvoeringsbesluite n_4.pdf (last consulted on 15/01/2016). 80 http://www.levif.be/actualite/belgique/les-regions-bruxelloise-et-flamande-s-appretent-a-legiferer- sur-Ai rbnb-etc/article-normal-405427.html (last consulted on 30/10/2015).

PE 558.777 II - 152

Germany, Spring 2015. This report states that in February 2015 Airbnb offered 11.700 apartments in Berlin. Munich (around 4000), Hamburg and Cologne (around 3000), came in second, third and fourth place. The report did not come to a clear conclusion on the consequences of subletting for rental prices in general.81

In order to meet general discomfort from the side of citizens and hotel organisations, the city of Berlin adopted a local law in May 2014: the so-called Zweckentfremdungsverbot-Gesetz (ZwVbG).82 The law bans the regular short-term letting of rooms without permission from the authorities. However, a two-year's term was included for apartments that were already being used for holiday lets at the time the law came into force.

Based on this law, at the end of 2015 approximately 6,300 apartments were registered and 1,200 possible violations. There are also around 2,800 notes from the public on possible infringements of the ban, which are also being checked.83 Inspections are conducted in order to ensure that the law is correctly applied, authorising compliance forces to enter into the houses without warrant84.

At the moment of writing this report (December 2015) a new draft law is being discussed – an amendment to the existing Zweckentfremdungsverbot-Gesetz (ZwVbG). It stipulates that secondary residences cannot be rented without a permit as a holiday home. In addition, according to the new draft law the authorities will be able to oblige platforms like Airbnb and similar ones like Wimdu and 9Flats to collaborate in the investigation of possible cases of infringement. That is, they will have to give the District Offices information about the registered users of their portals.85

France

In response to the creation of platforms such as Airbnb, the French legislator intervened mainly through the adoption of the so-called 'Loi ALUR'86. Occasional users wishing to rent a room see their rules simplified as they do not need to request an authorization. In this regard, a distinction is made between principal and secondary residences. Short-term rentals are allowed without requiring a prior authorization from the City Council for principal residences. If the dwelling is the owner's secondary residence, short-term rentals are also authorised, but in cities of more than 200.000 inhabitants the owner must obtains a prior authorization from the City Council87. Otherwise, short-term rentals regularly used are commercial units, subject to the

81 http://www.Airbnbvsberlin.com/. 82 http://gesetze.berlin.de/jportal/?quelle=jlink&query=WoZwEntfrG+BE&psml=bsbeprod.psml &max=t rue (last consulted on 10/01/2016). 83 http://www.gastronomie-hotellerie.com/berlin-verschaerft-kampf-gegen-Airbnb (last consulted on 10/01/2016). 84 http://www.spiegel.de/international/business/berlin-cracks-down-on-estimated-18-000- vacation-rentals-a-1026881.html (last consulted on 01/11/2015). 85 http://www.gastronomie-hotellerie.com/berlin-verschaerft-kampf-gegen-Airbnb (last consulted on 10/01/2016). 86 Loi n° 2014-366 du 24 mars 2014 pour l'accès au logement et un urbanisme rénové (« ALUR ») [Act n° 2014- 366 of 34 March 2014 on the access to housing and renewed urbanism]. 87 The other cities enjoy discretion to decide whether the prior obligation is compulsory or not.

PE 558.777 153

common legal and tax framework88. Violating these rules entails fines of up to 25000 Euros, and French tribunals effectively enforce these rules89.

Therefore, using Airbnb and other similar platforms is legal provided that the owner complies with the existing regulation, including tax legislation, which tends to align to the hotels' regime. For example, the French Government recently adopted a decree that extended the tourist tax to this situation90 and it is quite significant that Airbnb created a mechanism to collect this tax on behalf of the city council. This mechanism is now in force in Paris, but it will be progressively expanded to the other French cities91.

Nonetheless, the legislation is not completely clear, especially regarding the threshold upon which the dwelling becomes a commercial unit and using Airbnb becomes a professional activity. So far, no judgment has been published in this regard92. However, a tax exemption for revenues stemming from the SE amounting up to 5000 Euros is under discussion by the national authorities93. Above that threshold, the common legal and fiscal provisions would apply. In any event, it should be noted that the Act did not change anything regarding subletting. Therefore, a tenant cannot use Airbnb and other similar platforms to rent out the dwelling, except if the owner expressly consented to it and the rent does not exceed the one paid by the tenant94. Violating these rules may entail the lease cancellation. The first judgment that involved Airbnb took place in 2014: a tenant who hosted tourists through Airbnb was condemned, but the tribunal refused to authorise his eviction from the apartment95.

The Netherlands

The City of Amsterdam is one of the European cities with the highest Airbnb apartment's density. On a population of less than a million inhabitants, some 13.000 rooms/apartments are rented out, of which 7.000 by Airbnb.96

88 Code of construction and housing, article L. 631-7. 89 http://droitdupartage.com/2014/09/24/Airbnb-une-utilisation-plus-risquee-que-lon-ne-le-croit/ (last consulted on 28/10/2015). 90 Décret n° 2015-970 du 31 juillet 2015 relatif à la taxe de séjour et à la taxe de séjour forfaitaire, JORF n°0179 of 5 August 2015, p. 13416. 91 http://www.lefigaro.fr/secteur/high-tech/2015/08/24/32001-20150824ARTFIG00330-Airbnb-va- collecter-la-taxe-de-sejour-pour-le-compte-des-mairies-francaises.php (last consulted on 28/10/2015). 92 http://droitdupartage.com/2014/09/24/Airbnb-une-utilisation-plus-risquee-que-lon-ne-le-croit/ (last consulted on 28/10/2015). 93 Michel Bouvard, Thierry Carcenac, Jacques Chiron, Philippe Dallier, Jacques Genest, Bernards Lalande, Albéric de Montgolfier, Rapport d'information au nom de la commission des finances n° 690 (2014- 2015) sur 'L'économie collaborative : propositions pour une fiscalité simple, juste et efficace ». Sénat. 17 September 2015. Available at: http://www.senat.fr/notice-rapport/2014/r14-690-notice.html (last consulted on 08/01/2016). 94 http://droitdupartage.com/2014/09/24/Airbnb-une-utilisation-plus-risquee-que-lon-ne-le-croit/ (last consulted on 28/10/2015). 95 http://www.legalis.net/spip.php?page=jurisprudence-decision&id_article=4140 (last consulted on 28/10/2015). 96 http://zoeken.amsterdam.raadsinformatie.nl/cgi- bin/showdoc.cgi/action=view/id=234437/type=pdf/4_Bi jlage_3_Overzicht_websites.pdf (last consulted on 10/01/2016).

PE 558.777 II - 154

On February 14, 2014, the Amsterdam City Council created a new category of accommodation that makes it legal for city residents to occasionally rent their homes to tourists.

The new category 'private rental' gives residents the opportunity to list their homes on sites like Airbnb without fear of penalty. But some rules do apply:

− Residents can only rent the home they live in and they must own the space or have permission from their landlord to rent it. − Residents must pay tax on the income made from short-term rentals as well as a tourist tax. − No more than four people are allowed to rent one home at a time and residents cannot rent their home for more than four consecutive nights. − A maximum of 60 days/year. Renting homes more than 60 days/year will be considered as commercial exploitation. − Tourists may not cause any inconvenience and the house must meet fire safety requirements.97

With this new legislation, Amsterdam became the first European city to pass an 'Airbnb friendly law' to support the sharing economy under certain conditions. It was welcomed by Henk Kamp, the Minister of Economy: 'If we in the Netherlands want to be first to profit from the benefits of innovation, then we have to make room for that in our rules'.

In December 2014 Airbnb and the city of Amsterdam came to a Memorandum of Understanding (MOU): Airbnb offered its help to collect the tourist taxes (5% of accommodation price). Airbnb will report information about infringement of the rules set by the City, on demand. The City Council published neither the text of the memorandum nor the details of the collaboration on a day-to-day basis. Both parties agreed not to start legal procedures against the other during the trial period, which was set at one year. In December 2015 the cooperation was renewed for another year, waiting for a more definite evaluation in spring 2016.98

Since the introduction of the new regulation, the City Council seems to be satisfied with the implementation, although practical problems had to be overcome. Especially the controlling of more than 10.000 offerings was a more costly and complicated job than expected. In December 2015, the platform removed some 170 Amsterdam ads because the landlords would not abide by the rules99. More details about the effectiveness of the new legislation and the collaboration with Airbnb are expected to be published in the fall of 2016.

97 https://www.amsterdam.nl/veelgevraagd/?caseid=%7B9B2C2273-F797-460B-AD20-05DFB9F6F39F% 7D (last consulted on 10/01/2016). 98 https://www.amsterdam.nl/gemeente/college/nieuws-uit-b-w/2015/nieuws-16- december/#h41a09c67-2c 3f-46c4-a157-593e3f164285 (last consulted on 10/01/2016). 99 http://www.volkskrant.nl/economie/airbnb-schrapt-huizen-amsterdam-die-te-vaak-worden- verhuurd~a4 223391/ (last consulted on 10/01/2016).

PE 558.777 155

Meanwhile other Dutch cities, like Rotterdam, The Hague and Utrecht are thinking of similar agreements with Airbnb, but prefer to wait for the outcome of the 'Amsterdam experiment'.100 The efforts of Amsterdam to come to an agreement with other platforms than Airbnb, have not been fruitful so far.101

Spain

It should be noted that since 2013, the regulation of tourist accommodation is regulated at the regional level102. Several measures have been adopted in Spain in reaction to the emergence of Airbnb-like platforms, in particular in Catalonia.

Catalonia According to the Catalan tourist legislation103, hotels and tourist apartments are subject to a licence fee. Furthermore, it is prohibited to rent-out a single room in a private apartment104. In July 2014 Airbnb was one of the eight letting sites fined by the Catalan government (30.000 Euro fine) for a 'serious infringement' of the legislation105. The long battle against private lets is not specifically against Airbnb, but it has been caught in the pack.

While hotel owners claim unfair competition of the site, neighbourhood associations blame private lets for driving up house prices in central districts, plus putting locals in the position of being neighbours of an ever-changing roster of tourists, causing the effect of all times neighbours moving out to less touristic areas.

As a consequence the Catalan Government (Generalitat) is preparing a set of rules that apartment owners will have to comply with in order to put their rooms and apartments in the market106. The proposal is not public yet but its main lines have been already presented to the press. Owners will be allowed to rent apartment rooms under the following conditions:

− They live in the apartment, before and during the rental period.

100 http://nos.nl/op3/artikel/2076801-den-haag-rotterdam-en-utrecht-praten-met-Airbnb-over- toeristenbelast ing.html (last consulted on 10/01/2016). 101 http://zoeken.amsterdam.raadsinformatie.nl/cgi- bin/showdoc.cgi/action=view/id=234434/type=pdf/1_R apportage_vakantieverhuur.pdf (last consulted on 10/01/2016). 102 Ley 4/2013, de 4 de junio, de medidas de flexibilización y fomento del mercado del alquiler de viviendas. 103 The most important acts are the following: Ley 13/2002, de 21 de junio, de turismo de Cataluña, DOGC nº3739, 14/10/2002; ley 18/2007, de 28 de diciembre, del derecho a la vivienda, DOGC nº5044, 09/01/2008; Decreto 159/2012, de 20 de noviembre, de establecimientos de alojamiento turístico y de viviendas de uso turístico, DOGC nº 6268, 05/12/2012. See: http://empresaiocupacio.gencat.cat/es/treb_departament/emo_normativa/emo_normativa_turisme/ (last consulted on 08/01/2016). 104 Decreto 159/2012, de 20 de noviembre, de establecimientos de alojamiento turístico y de viviendas de uso turístico, DOGC nº 6268, 05/12/2012, article 66(2). 105 http://premsa.gencat.cat/pres_fsvp/AppJava/notapremsavw/273437/ca/generalitat-sanciona- Airbnb-comercialitzar-habitacions-domicilis-particulars-oferta-illegal-normativa-garantia-qualitat.do (last consulted on 08/01/2016). 106 http://novobrief.com/Airbnb-legal-in-cataluna-barcelona/ (last consulted on 10/01/2016).

PE 558.777 II - 156

− Rentals cannot last more than 31 days and rooms can only be available for a maximum period of 4 months per year (not consecutive). − Maximum of 2 rooms per apartment. − Municipalities will be able to determine in which city areas this activity can take place. − Owners will be responsible of collecting a tourist tax (€0.65 in Barcelona and €0.45 in the rest of Catalonia, per night).

In the case of owners who wish to rent whole apartments, the Catalan Government will require them to include their properties on Catalonia's tourism registry and to have a tourist license.

Besides, Barcelona local authorities have implemented strict measures in order to detect illegal apartments and they imposed fines of 60.000 Euros to Airbnb in December 2015107. The authorities denounce the fact that Airbnb advertised rooms without their registration number, in violation of the Catalan legislation. Airbnb already announced that it would challenge the decision. It is worth noting that Barcelona has stopped in the summer 2015 issuing new tourist licenses for hotels and holiday rentals for at least one year108.

Madrid Madrid has yet to pass a law regulating Airbnb's activities in the Spanish capital. In 2014 it said it would impose a 5-day minimum stay for apartment rentals, but still today one can easily find shorter rentals on the site.

United Kingdom109

In February 2015 the Department for Communities and Local Government, UK Government, decided to review and modernise Property conditions in the private rented sector for London. While in all other parts of the country residents are able to let out their homes for short periods as a matter of course, in London short-term use is strictly regulated under legislation dating back to the 1970s110.

London had enacted legal provisions in order to protect London's existing housing supply, for the benefit of permanent residents, by giving London boroughs greater and easier means of planning control to prevent the conversion of family homes into short term rentals111.

Short-term use as temporary sleeping accommodation was only permitted once planning permission is obtained from the local authority. London residents faced a possible fine of up to £20,000 for each 'offence' of failing to secure planning permission. This measure was preventing Londoners from participating in the SE by letting out either a spare room or their whole house in the same way as other residents across the country.

107 http://ccaa.elpais.com/ccaa/2015/12/21/catalunya/1450694300_272687.html (last consulted on 08/01/2016). 108 http://ccaa.elpais.com/ccaa/2015/07/02/catalunya/1435818964_383457.html (last consulted on 12/01/2015). 109 www.Airbnb.co.uk/help/responsible-hosting (last consulted on 25/10/2015). 110 Section 25 of the Greater London Council (General Powers) Act 1973. 111 UK Government - Deregulation Act 2015, Sections 44 and 45: Short-term use of London accommodation: relaxation of restrictions and power to relax restrictions. Available at: www.legislation.gov.uk/ukpga/2015/20/notes/division/5/46 (last consulted on 25/10/2015).

PE 558.777 157

Adoption of legislation favouring the SE

In 2015, with the Deregulation Act, the Government introduced an exception to this restriction.

'London is a great city, and the Government wants to give Londoners the opportunity to be part of this modern approach and able to participate fully in the sharing economy. We want to open up this great global city to more visitors by embracing these new opportunities to allow people to make better use of their property'112.

The exception allows residential premises to be used for temporary sleeping accommodation without this being considered a 'change of use', so long as the cumulative number of nights of use as temporary sleeping accommodation does not exceed 90 nights in a calendar year, and so long as the person who provides the accommodation is liable to pay council tax.

'Local planning authorities may direct that this exception does not apply to certain residential premises or to residential premises in certain areas. If a property is used for short-term rentals for more than 90 days in a calendar year, the exception does not apply.' The Government clarifies that this policy is aimed at helping residents, and not providing opportunities for the commercial sector.

Taxes

The UK Government has established 'The Rent a Room Scheme'113 allowing residents earn up to a threshold of £4,250 per year tax-free from letting out furnished accommodation in their home. This is halved if they share the income with their partner or someone else.

The problem of property permanently in short-term use

'It was clear that the existing focus for concern was where a property was permanently in short- term use, and this was where enforcement actions were largely targeted. The need for other related regulations, such as health and safety legislation, to remain in place was also recognised as important'114.

Local authorities also confirmed that in considering any enforcement they would usually be mindful that some forms of short-term use will already be classed as a material change of use under Building Regulations, which results in the need for work to be carried out to upgrade the building to appropriate standards. There are also Fire Safety provisions that would need to be checked for compliance.

112 Department for Communities and Local government, Promoting the sharing economy in London, Policy on short-term use of residential property in London, 2015. Available at: www.gov.uk/government/uploads/system/uploads/attachment_data/file/402411/Promoting_the_s haring_economy_in_London.pdf (last consulted on 25/10/2015). 113 www.gov.uk/rent-room-in-your-home/the-rent-a-room-scheme (last consulted on 25/10/2015). 114 Department for Communities and Local government, Promoting the sharing economy in London, Policy on short-term use of residential property in London, 2015, para. 20. Available at: www.gov.uk/government/uploads/system/uploads/attachment_data/file/402411/Promoting_the_s haring_economy_in_London.pdf (last consulted on 25/10/2015).

PE 558.777 II - 158

2.3. Professional services

This section covers a wide array of SE platforms offering services ranging from postal services to freelance work for which otherwise it would be necessary to hire a professional (cleaning, accounting, designing, etc.). Because of the all-encompassing nature of this type of service provision we will first provide an introduction before discussing the benefits and complaints and presenting the national responses.

Service provision through platforms is not new, but 'a whole slew of labour platforms have come up over the last couple of years, powering what is widely referred to as the sharing economy, Platform economy or (called by many) the Gig Economy'115.

According to the book 'Platform Scale: How an emerging business model helps start-ups build large empires with minimum investment'116remote freelancing (Freelancer, Elance-Odesk) and micro-tasking (Amazon Mechanical Turk) platforms have been around for quite some time, all of which enable service providers to find new 'job' (gig) opportunities. But a whole new range of specific task-related platforms have come up in recent times creating two broad classes of new opportunities:

1. Higher end jobs: Consulting platforms like Clarity and Experfy now enable highly skilled individuals to find (temporary) jobs on platforms. 2. Real world job coordination: Platforms like Homejoy and Postmates allow people with spare time to find a new source of income in the 'real' world. 2.3.1. Benefits

The benefits for consumers are the same as with SE platforms in the transport and accommodation sector. (Sometimes luxury) services (personal assistants, cleaning services) become reasonably priced and affordable to much more people than before. Consumers can enjoy a lifestyle that was previously unthinkable. The professional services are on-demand, whereas traditional service providers often come with waiting periods.

2.3.2. Complaints

Professional service (or labour) platforms have been welcomed with great enthusiasm, but time is proving that they cannot be conceived as an alternative to traditional jobs. Many important aspects of traditional work are not offered in the existing online labour platforms (stability, healthcare, pension, paid holidays, social contacts etc.). Most complaints against professional service companies come from their own users. A number of people that use the platforms to earn a living complain about their employment status. Even though the platforms impose certain rules on them (e.g. what to wear, what products to use, etc.), workers are still considered as independent contractors instead of employees. Workers providing services through labour platforms are not enjoying the benefits of traditional contract employment.

115 http://platformed.info/sharing-economy-future-of-work/ (last consulted on 10/01/2016). 116 CHOUDARY, S.P. (2015) Platform Scale: How an emerging business model helps start-ups build large empires with minimum investment. Platform Thinking Labs.

PE 558.777 159

2.3.3. Trends

The platform owners are moving beyond the pure role of matching the supply and demand side, and start to understand and deal with the costs and benefits involved for the labour force participating on these platforms.

Unlike traditional organisations, workers may not be contractually tied to only one platform; most workers will participate in several platforms. This is why if platforms offer a type of insurance, it is often only partial, because they are reluctant to cover activities that take place beyond the platforms' sphere. Therefore, in order to guarantee the well being of their users, platforms will need to develop horizontal infrastructural services.

According to the book cited previously - 'Platform Scale: How an emerging business model helps start-ups build large empires with minimum investment' by S.P. Choudary -, these horizontal services are not yet put in place. This is the case for – for example - traditional insurance, where they have not yet developed horizontal insurance products addressed to service providers participating in multiple platforms (also called 'multi-homing'). The freelancer economy may need an insurance provider specifically suited for this purpose. In some other cases, timid gestures are already performed, showing an initial emergence of such horizontal initiatives. Peers.org, for example, is powering the counterpart of labour unions for the platform economy. Sherpasphere and other similar players are powering better job discovery and management. Nevertheless, affordable healthcare guarantees for a platform economy continue to be elusive. And finally, as jobs get unbundled even further, these multi-homing service providers will need greater back office support. Most freelancers on labour platforms today are contract employees. They have unique work management and taxation issues. Services like 1099.is, TryZen99 and UseBenny have emerged to address this but there are many more opportunities to provide greater infrastructural support to this emerging economy117. In the words of S. P. Choudary, 'The future of job creation isn't just about matching supply to demand but about providing the entire infrastructure that enables producers to reliably find a better substitute than traditional job alternatives. To enable this, platforms should ensure favourable producer participation economics. In particular, non-platform players need to emerge to provide infrastructural services across multiple platforms'118.

2.3.4. Case analysis

United States of America

Just like Uber and Lyft119, on-demand platforms for services in other sectors have also been hit by lawsuits regarding the question whether their users are independent contractors or employees.120 Employees generally have certain rights, such as overtime, a minimum salary,

117 For more information on the networked platforms (multi-homing) see Section 3.4 'The creation of cumulative value' in the book by CHOUDARY, S.P. (2015) op. cit. 118 Ibid., p. 80. 119 https://www.lyft.com/ Operating only in USA. 120 http://www.bizjournals.com/sanfrancisco/blog/2015/03/homejoy-independent-contractors-uber- lyft-lawsuits.html (last consulted on 10/01/2016).

PE 558.777 II - 160

workers compensation (replacement wage and medical benefits during medical leave), etc. These are traditionally not granted to independent contractors.

The service providers in these platforms (drivers, cleaners) basically want to have the choice between having more rights as a full employee or otherwise having the possibility to exert more control over their conditions so they can, for instance, share a higher hourly rate.121

All the cases that have been filed so far took place in the US. The claims are pretty similar: users should receive the same benefits as employees given their treatment. This treatment differs from platform to platform, but each of them submits their users (service providers) to certain rules/controls which may indicate that they are employees under law. Below, we provide an overview that is taken from Carson's (2015) article122.

The on-demand laundry and dry cleaning service company, Washio, hires drivers, or 'ninja's', to deliver laundry to their customers. The company is said to make drivers sign an exclusivity agreement, where they agree not to work for similar businesses. According to the court filing, drivers are paid a fee for each pick-up and delivery.

As a provider of similar services, Shyp choses to pay its drivers an hourly fee to pick up packages (vs. a fee per pick up). Drivers are given clear instructions on how to perform their tasks, for instance: fragile items have to be bubble wrapped. In addition, drivers claim to receive warnings if they reject more pick-ups than allowed.

Postmates was also hit by a court case. According to the arbitration demand, Postmates' service providers only make $0, 35 per delivery. In addition, the company is said to impose the following rule on its service providers: if a store does not answer the phone, one has to call back four times in two minutes before launching a search to find out if the store is closed. One plaintiff estimated that in April 2015, she worked thirty hours but only made $45, 85 (approximately €42) for 131 calls.

Table nº 3 Platforms' rules and controls on service providers

Company 'Job' conditions Non-competition clause imposed on service providers + commission paid to Washio the platform Control over providers' performance + strict instructions about how to perform Shyp the job Postmates Strict instructions about how to perform the job

A case had also been filed against Homejoy, a platform for cleaning services, which is said to have led to the shutdown of the company.123

121 http://arstechnica.com/tech-policy/2015/03/startup-workers-sue-to-be-recognized-as-employees- not-mere-contractors/ (last consulted on 10/01/2016). 122 http://www.businessinsider.com/postmates-shyp-and-washio-hit-with-legal-action-from-contractors- 2015-7#ixzz3gopeDRKp (last consulted on 12/01/2016). 123 http://www.forbes.com/sites/ellenhuet/2015/07/17/cleaning-startup-homejoy-shuts-down-citing- worker-misclassification-lawsuits/ (last consulted on 10/01/2016).

PE 558.777 161

2.4. Comparative analysis of national legal responses

SE platforms are not easily reconcilable with pre-existing national legislation, in particular with regard to services like Uber, which severely affect established legal monopolies. The legal responses vary across Member States, and also depending on the type of SE platform at stake. Generally speaking, there is a shared concern by all, which is the necessity to adapt to technological innovations while ensuring the respect for fair competition. Nevertheless, this balance has not yet been found, as the varying responses demonstrate.

In the absence of cases on professional services found in Europe, the following section will focus on transportation and accommodation.

2.4.1. Complete banning and equalising to traditional services

Banning

Firstly, one of the responses has been banning. In some MS, the approach is quite conservative due to important mobilisation either from the traditional actors, or from the civil society.

This is particularly the case when talking about UberPOP, which has been banned in several countries or cities (DE, ES, FR, NL, Brussels) because it would not comply with the existing taxi legislation. The justifications for these bans make think that generally speaking, the national authorities and jurisdictions consider that UberPOP is a regular transport company which has been operating without the necessary licences. In the accommodation sector, the situation is a little bit more nuanced, insofar as unregistered short-term rentals are prohibited in some places if they are not duly registered (Berlin and Catalonia124).

Applying strictly the existing legislation to the SE

In some other cases, the legislation does not differentiate users of SE platforms from traditional actors. Even though we are not strictly talking about an intervention from the legislature, the decision is to apply the common legal framework to SE users, with all its consequences. For example, the other forms of Uber different from UberPOP (e.g., UberTaxi) are allowed in different MS as long as they abide by the normal rules applicable to taxi services, including licencing, liability and fiscal rules (DE, FR, NL and Brussels).

In the accommodation sector, we find examples in Berlin and Catalonia. Registered rentals, which comply with the existing tourism legislation, can perfectly be advertised on Airbnb-like platforms. However, all unregistered short-term rentals are prohibited, and in the case of Catalonia, single rooms in private apartments cannot be rented out. Due to the strong social mobilisation against short term rental in different parts of Catalonia, there is a strong willingness to enforce the existing legislation very strictly: The Catalan Tourist Act was recently

124 A bill is being discussed to change the situation in Catalonia, but for now, the current tourism legislation forbids to rent out single rooms in private apartments.

PE 558.777 II - 162

modified so as to oblige third parties to advertise the registration number of holiday rentals in any type of advertisement in which they appear125. This measure was approved in order to provide the administration with more efficient means to control illegal rentals.

In this respect, some public authorities have taken seriously the need to police not only users (in the sense of providers), but also platforms. The Catalan case is a good example. The Generalitat and Barcelona local authorities respectively fined Airbnb with 30.000 Euros in July 2014, and 60.000 Euros in December 2015. It is worth mentioning that the fines were imposed because Airbnb allegedly advertises rentals that are not registered, and are therefore illegal.

Therefore, some aspects are legalized but there is an alignment of the legislation with the regime that applies to traditional actors. The underlying idea is that these new platforms only take advantage of technological advances to provide already existing services. The new actors are qualified as commercial actors that must comply with the same rules as those that apply to the traditional actors.

The effects of this approach by the authorities is starting to have results: In Amsterdam, Airbnb is up to fight illegal hotels promoted on its platform. The houses rental site removed in late December 2015 some 170 Amsterdam ads because the landlords would not abide by the rules.126

2.4.2. Imposing more rules

In many cases, the MS' response has been to regulate the activity at stake, but with measures that impose more rules.

An example from the transport sector that arguably hinders the SE is the prohibition on electronic cruising approved in France. In the same way, even though they have not been adopted yet, the different measures that are being discussed in the UK are likely to have a negative impact on Uber-like platforms (e.g., a ban on showing cars for hire within a smartphone app, the ability to book rides up to seven days in advance, or the creation of controls on ridesharing in public vehicles).

In the accommodation sector, with the obligation to obtain a registration number for single rooms, Brussels' response has been to impose more rules, obligations and controls on Airbnb's users. The same holds true in Berlin where unregistered short-term rentals are expressly prohibited. In Barcelona, the new draft legislation is going to impose strict obligations on accommodation owners.

125 Ley 2/2014, de 27 de enero, de medidas fiscales, administrativas, financieras y del sector público, DOGC n.6551, 30/01/2014. 126 http://www.volkskrant.nl/economie/airbnb-schrapt-huizen-amsterdam-die-te-vaak-worden- verhuurd ~a4223391/(last consulted on 10/01/2016).

PE 558.777 163

2.4.3. Simplifying

In the transportation sector, the tendency has not been to simplify the rules. However, one element is worth mentioning: the Loi Thévenoud expressly allows car sharing platforms that follow the model of Blablacar, i.e., where passengers share the costs of the ride with the driver, and where the trip is carried out by the driver for their own purposes.

Some authorities have revised the existing rules to make them more SE-friendly (Flanders, Hamburg, FR, NL, UK) in the accommodation sector even though they remain subject to a certain number of conditions.

In this regard, Flanders, FR, Hamburg, NL and Wallonia allow short-term rentals without licence or prior authorisation if the owner lives in the residence. The restrictions are therefore imposed on the secondary residence of the owner, which is subject to a licence or prior authorisation. Yet, in the French case, this obligation is automatic only in cities of more than 200.000 inhabitants.

In the field of tax collection, tools have been found in some cases to facilitate compliance with the law, through collaboration between the SE platform and the public authorities. The creation of a system of collection of the tourist tax directly on Airbnb is a good example (FR and NL). In the same way, France is currently discussing the possibility of creating an automatic system of tax return, in the form of a pre-filled tax return directly available on SE platforms127. However, in order not to establish a too burdensome system for occasional users, it is also foreseen to grant a tax exemption for revenues amounting up to 5000 Euros; above that threshold, the common legal and fiscal provisions would apply. Another measure that is currently being discussed is the possibility to collect VAT directly at the source, at the moment of the transaction128. This measure is not foreseen for SE platforms only, but for all eCommerce services. The idea is not to add a new tax, but to guarantee the effective recovery of this tax which is due, but not collected.

Therefore, the existing legal provisions in this domain are quite sparse; A common legal and tax framework covering the different activities falling under the SE is therefore needed. Nevertheless, the existence of scattered legislation also demonstrates that reaching an agreement at EU level might prove to be difficult.

2.4.4. Best practices

127 Michel Bouvard, Thierry Carcenac, Jacques Chiron, Philippe Dallier, Jacques Genest, Bernards Lalande, Albéric de Montgolfier, Rapport d'information au nom de la commission des finances n° 690 (2014- 2015) sur 'L'économie collaborative : propositions pour une fiscalité simple, juste et efficace ». Sénat. 17 September 2015. Available at: http://www.senat.fr/notice-rapport/2014/r14-690-notice.html (last consulted on 08/01/2016). 128 Michel Bouvard, Thierry Carcenac, Jacques Chiron, Philippe Dallier, Jacques Genest, Bernards Lalande, Albéric de Montgolfier, Rapport d'information au nom de la commission des finances n° 691 (2014- 2015) sur ' Le e-commerce : propositions pour une TVA payée à la source». Sénat. 17 September 2015. Available at: http://www.senat.fr/notice-rapport/2014/r14-691-notice.html (last consulted on 08/01/2016).

PE 558.777 II - 164

Some Member States have adopted measures that are interesting to foster the SE and which could be exported to other cities or MS. Table nº 4 briefly describes them.

Table nº 4: Best practices

Area Best practices

The UK SE sector is engaging in associations for their sector in order to early engage with the regulators to design common constructive solutions for the P2P business. Decision- making and Brussels authorities are discussing a new legal framework for alternative taxi services procedures in consultation with the taxi industry. This practice is especially relevant in this sector where taxis benefit from monopolies and are able to mobilize themselves in an important manner.

Providing a clear definition of the key concepts of the SE is crucial to ensure that citizens and businesses correctly apply the law. In this regard, the clarification on the concept of car sharing in France, based on Blablacar's model, is welcomed. Concepts and In addition, the Dutch and British authorities provide clear criteria to distinguish a definitions commercial activity from occasional use, including time and space limits, safety requirements, and questions of ownership (residents can only rent the home they live in and they must own the space or have permission from their landlord to rent it). It therefore allows citizens and businesses to know very precisely if and how they can legally make use of Airbnb-like platforms, without room for doubt.

Cutting red tape is extremely important for the success of the SE. The legislation in Flanders and in Wallonia is efficient in cutting red tape in the accommodation sector. In particular, the ease of becoming an official touristic accommodation operator in Cutting red Flanders, by voluntarily requesting a recognition is especially worth noticing. tape Furthermore, one aspect of the legislation in Brussels is worth mentioning: while a normal procedure for getting a certificate of fire safety requires an inspection from the fire department, individuals will only need to present an inspection certificate of gas and electricity.

Cooperation with SE platforms to ensure compliance with the law is essential. In this regard, the system of collection of the tourist tax by Airbnb on behalf of public authorities (FR, NL) is extremely interesting and could be extended to other cities. Tax compliance In the same way, the British authorities grant tax exemptions on revenue amounting up to £4,250 per year for letting out furnished accommodation in their home. In France, the authorities are discussing a similar system that would apply to SE platforms in general, and which would amount to up to 5000 Euros. That kind of arrangement is important to foster citizens' use of SE platforms without fearing of infringing the law. It is also important to ensure that the benefits outweigh the costs of using SE platforms

PE 558.777 165

Even though these measures are under discussion before the French authorities, it is worth to mention: - The proposal to create an automatic system of tax return, in the form of a pre- filled tax return directly available on SE platforms; and - The proposal to collect VAT directly at the source, at the moment of the transaction. This measure is not limited to SE platforms only, but to all digital services in general.

Cooperation between the SE platforms and the public authorities is probably key to the enforcement of the legislation. In this respect, the system established between Amsterdam and Airbnb, according to which Airbnb will report information about infringement of the rules set by the City upon request, could serve as a model for other Enforcement cities. Instead of adopting a punitive approach towards the platform for publishing illegal rentals, cooperation is foreseen to identify them.

The software used in the City of Barcelona to detect illegal rentals is also worth mentioning as it probably is more efficient and less time consuming than the systems such as the one established in Berlin.

PE 558.777 II - 166

3. The EU regulatory framework

The SE is high on the EU agenda and its impact on the European single market is currently being scrutinized. In particular, the European Commission wishes to adopt EU legislation in this regard, in order to provide a unified response to the fragmented domestic legal frameworks. According to Commissioner Elżbieta Bieńkowska, the idea is 'to present new rules this year to regulate and enable cross-border digital services such as Uber, Airbnb and other new online businesses to work more smoothly in the Member States'129.

In this regard, the Commission has referred to the benefits of the SE in its Communication on a Digital Single Market Strategy for Europe (DSM) of 6 May 2015130, where it decided to assess 'the role of platforms, including in the sharing economy, and of online intermediaries'. As part of this assessment, it has launched an online public consultation from 24 September till 30 December 2015131 monitored by DG for Communication networks, Content and Technology, and DG Internal Market, Industry Entrepreneurship and SMEs. In parallel, the Commission launched two studies, one by the Directorate-General for Mobility and Transport on 'passenger transport by taxi, hire car and ridesharing in the EU' and another one by the Directorate- General for Justice and Consumers on 'consumer issues in the sharing economy', both expected by the second quarter of 2016.

More recently, in the Communication on Upgrading the Single Market: more opportunities for people and business of 28 October 2015132, the Commission said that it would provide guidance on how EU law applies to collaborative economy business models, rather than strictly regulating the issue in 2016. In particular, it will draw upon national, European and international existing legislation to identify best practices, analyse how regulatory gaps need to be filled, and monitor its development. Thus, while the results of these initiatives are still to be expected, some elements can already be discussed.

But first of all, the question regarding whether the EU holds competence to act on this matter needs to be raised. The answer to this question can be answered positively: The Treaty on the Functioning of the European Union (TFEU) gives competence to the EU in 'the establishing of the competition rules necessary for the functioning of the internal market' (article 3). As a consequence, if one considers the objectives which are pursued by the Single Market regulatory framework133, legal action at EU level seems necessary in the SE field, in particular with regard to the necessity to remove existing barriers to intra-EU trade and preventing the creation of new ones.

129 https://global.handelsblatt.com/edition/204/ressort/politics/article/removing-europes-trade- hurdles (last consulted on 01/11/2015). 130 COM(2015) 192 final. Available at: http://ec.europa.eu/priorities/digital-single-market/docs/dsm- communication_en.pdf (last consulted on 02/11/2015). 131 https://ec.europa.eu/digital-agenda/en/news/public-consultation-regulatory-environment- platforms-online-intermediaries-data-and-cloud (last consulted on 01/11/2015). 132 COM(2015) 550 final. 133 The objectives are the following: free movement of services; necessity to remove existing barriers to intra-EU trade and preventing the creation of new ones; and promoting a business and consumer- friendly environment based on transparent, simple, and consistent rules offering legal certainty and clarity. See: http://ec.europa.eu/growth/single-market/index_en.htm (last consulted on 01/11/2015).

PE 558.777 167

Furthermore, when considering the different types of SE platforms analysed in this report, there is already an important corpus of legislation that could serve as a basis for further action on this matter, albeit the EU action is unevenly important.

Transport is a shared competence between the EU and the Member States (TFEU, article 91); however it may apply to some SE platforms only, as it must involve a transnational aspect. Therefore, it seems that it would apply to Blablacar-like platforms rather than Uber-like platforms.

Moreover, the accommodation sector can be linked to tourism134, where Union action must complement the action of the Member States but cannot lead to harmonization. Consequently, the degree of EU action will vary depending on the type of service provided.

Finally, many of the legal issues raised in Susan Mclean's (2015) article135 are at least partly regulated at EU level. In this respect, there is a set of EU legislation that ensures consumer protection136, data protection137, protection from discrimination138, and which deals with employment matters139 at EU primary and secondary levels.

The amount of European legislation affected directly or indirectly by the SE is therefore very vast. However, due to time and budget limitations, we will concentrate specifically on services, eCommerce, consumer protection and data protection in order to analyse to what extent are the current rules fit for purpose, and whether or not there is a need for the European Union to legislate and to what extent. Indeed, those areas are probably the most relevant ones at EU level in the SE field at this stage. The European Commission referred to them in the Communication on Upgrading the Single Market: more opportunities for people and business of 28 October 2015, said that they would be scrutinised and, if needed, modified.

Services Directive

Most of the services offered by the SE platforms probably meet the requirements of article 57 TFEU, which defines them as follows: 'Services shall be considered to be 'services' within the meaning of the Treaties where they are normally provided for remuneration, in so far as they are not governed by the provisions relating to freedom of movement for goods, capital and persons'.

Directive 2006/123/EC on services in the internal market (hereafter, the 'Service Directive') provides further guidance on its meaning. The Service Directive aims to guarantee the effective exercise of the freedom of establishment for providers and the freedom of provision of services between Member States as established by articles 49 and 56 TFEU. It covers most of regulated professions and tourism, with the exception of transport services. In particular, it applies when

134 TFEU, articles 6 and 149. 135 MCLEAN, S., (2015) 'The rise of the sharing economy', Computers and Law magazine, Vol 26 Issue 1 April/May 2015. 136 Charter of Fundamental Rights of the EU, article 38; TFEU, articles 4 and 169. Directive on Consumer Rights (2011/83/EC). 137 Charter of Fundamental Rights of the EU, article 8. 138 Charter of Fundamental Rights of the EU, article 21; TFEU, articles 10, 18, and 19. 139 TFEU, articles 2, 5, 9, 45-48.

PE 558.777 II - 168

a) an undertaking wishes to establish itself permanently in its own country or in another Member State, and b) in case of cross-border service provision, in particular, when a consumer wishes to receive a service from another Member State, or when an undertaking established in a Member State wishes to provide services in another Member State without establishing itself permanently there. It is therefore critical to the good functioning of the single market, including with regard to the SE.

eCommerce Directive

Directive 2000/31/EC on electronic commerce (hereafter, the 'eCommerce Directive') establishes the legal framework for information society services (or online services) in the single market. Its objective is to remove 'obstacles to cross-border online services in the European Union and provide legal certainty to business and citizens in cross-border online transactions'140.

According to the current legislation, information society services can be defined in the following terms: 'undertaking operating in two (or multi)-sided markets, which use the Internet to enable interactions between two or more distinct but interdependent groups of users so as to generate value for at least one of the groups. Certain platforms also qualify as Intermediary service providers'141.

While the eCommerce Directive establishes a general duty not to restrict the freedom to provide information society services from another Member State (article 3), it allows them to derogate from this duty where the derogation fulfils objectives of public policy, protection of public health, public security and the protection of consumers. In this sense, the bans that have been established in different Member States, especially regarding UberPOP probably fulfil one or several objectives. As a result, this Directive is critical to assess whether the EU regulatory framework is fit for purpose.

The eCommerce Directive offers some sort of liability protection for platforms, as it establishes a liability exemption affecting contractual liability, administrative liability, tortuous / extra- contractual liability, penal liability, civil liability or any other of liability for the conduct of their users, provided that they fulfil Section 4 requirements142. This exemption includes conducts such as unfair commercial practices, unfair competition or publications of illegal content. Consequently, the Directive creates specific rules for information society services, thus establishing a difference of legal framework depending on whether the service is provided online or offline.

140 https://ec.europa.eu/digital-agenda/en/e-commerce-directive (last consulted on 19/12/2015). 141 Public consultation on the regulatory environment for platforms, online intermediaries, data and cloud computing and the collaborative economy, available at: file:///C:/Users/1236354/Downloads/pdfversionofthequestionnaire.pdf (last consulted on 12/01/2015). 142 EUROPEAN COMMISSION (2009) EU study on the Legal analysis of a Single Market for an Information Society. New rules for a new age? 6. Liability of online intermediaries. Brussels: European Commission, p. 9.

PE 558.777 169

Consumer legislation

While a number of concerns related to the sharing economy are linked with taxation and unfair competition issues, others are more consumer-centric. Concretely, we have briefly touched upon concerns related to insurance, discrimination, safety and security, liability and more broadly, consumer protection. It is therefore important to look at steps taken at the EU level on the matters that can prove useful in the area of the sharing economy.

First and foremost, the Directive on Consumer Rights143 has to be noted. With this Directive, the EU aims at 'at achieving a real business-to-consumer (B2C) internal market, striking the right balance between a high level of consumer protection and the competitiveness of enterprises.'144 Member States were to apply the national implementing laws as from June 13, 2014.

First of all, it lays down a number of information requirements. The 'trader'145 has to provide the main characteristics of the goods or services, reveal his identity, address and telephone number, present the total price, inclusive of taxes, as well as arrangements for payment and delivery. He also has to inform the consumer about his complaint handling policy as well as any relevant interoperability of digital content with hardware and software that the trader is aware of or can reasonably be expected to have been aware of.

Second, in article 20 the Directive clarifies who bears the risk at what moment when buying or service. It lays down that in contracts for the dispatching of goods form trader to consumer, the risk for loss or damage to the good passes to the consumer from the moment he (or a third party indicated by the consumer) has acquired the good in physical possession. However, if the consumer wants a carrier to transfer the good from the trader to the consumer and this option was not provided for by the trader, then the consumer bears the risk, without prejudice to the rights of the consumer against the carrier.

Finally, by article 22, it is prohibited for traders to use default options that the consumer needs to rejects in order to avoid additional payments. Instead, the trader should seek the express consent of the consumer to any additional payment.

However, with regards to consumer protection in the sharing economy, it is unsure whether the directive would be applicable to all types of SE platforms. While the directive applies to both sales and service contracts146, some areas are out of scope. To be sure, passenger transport services are not included nor is the 'supply of foodstuffs, beverages or other goods intended for current consumption in the household, and which are physically supplied by a trader on frequent and regular rounds to the consumer's home, residence or workplace'147.

143 Directive 2011/83/EU. 144 http://ec.europa.eu/consumers/consumer_rights/rights-contracts/directive/index_en.htm (last consulted on 10/01/2016). 145 ''trader' means any natural person or any legal person, irrespective of whether privately or publicly owned, who is acting, including through any other person acting in his name or on his behalf, for purposes relating to his trade, business, craft or profession in relation to contracts covered by this Directive;' (Directive 2011/83/EU, art. 2 (2)). 146 'service contract' means any contract other than a sales contract under which the trader supplies or undertakes to supply a service to the consumer and the consumer pays or undertakes to pay the price thereof; (Directive 2011/83/EU, art. 2 (6)). 147 Directive 2011/83/EU, art. 3 (3).

PE 558.777 II - 170

The Directive on Consumer Sales and Guarantees148 imposes rules on sellers of consumer goods with regard to the guarantee of conformity of the product with the contract. Since the entry into force consumers are guaranteed a period of two years after the delivery of the good. If, within this period, the good is not conform the sales contract, consumers have certain rights, for instance to see the good repaired or replaced. The final seller, on his part, can also hold the producer liable. This Directive only applies to the sales of consumer goods, understood as 'any tangible movable item'.

Also worth noting in this regard is the Directive on Unfair Contract Terms149. This directive aims to introduce a notion of 'good faith' so to prevent substantial imbalances between the rights and obligations of consumers on the one hand and sellers and suppliers on the other hand. Consumers can make use of the list of examples of terms that are considered unfair if they wish to do so, since if found unfair, terms are not binding for consumers. It furthermore stipulates that contract terms have to be written in plain language and that ambiguities are interpreted in favour of consumers.

Equally relevant is the Directive on Unfair Commercial Practices150, adopted in 2005, that aims to install fair commercial practices, by prohibiting the use of aggressive marketing techniques or by providing untruthful information. It covers the activities related to the promotion, sale and supply of both goods and services to consumers. It is further complemented by the Directive on Misleading and Comparative Advertising151, which establishes the criteria by which comparative advertisement is allowed and by the Price Indication Directive152 that stipulates that price information must be unambiguous, clearly legible and easily identifiable.

Finally, two more EU legal documents are worth discussing. The Alternative Dispute Resolution Directive153 provides consumers with an alternative to bringing a case to court when they have a problem with a trader regarding a good or service they purchased. Such an alternative can come in many forms and names, e.g. arbitration, mediation, ombudsmen. The directive ensures access to Alternative Dispute Resolution (ADR) no matter what good or service has been bought, irrespective they are bought online or offline. In addition to this directive, and under the Online Dispute Resolution Regulation154, the European Commission is establishing a European Online Dispute Resolution platform (ODR platform). This web-based platform is intended to help consumers that have problems with online purchases of goods or services. Available in all EU official languages, it allows citizens to submit their dispute and have their claim transmitted to a national ADR body.

Data protection legislation

Data protection rules were recently modified. On 15 December 2015, the European Parliament and the Council reached an agreement on the Data Protection Reform. The relevant instrument

148 Directive 1999/44/EC. 149 Directive 93/13/EEC. 150 Directive 2005/29/EC. 151 Directive 2006/114/EC. 152 Directive 98/6/EC. 153 Directive 2013/11/EU. 154 Regulation 524/2013/EU.

PE 558.777 171

regarding the SE is the 'General Data Protection Regulation' which aims to a) enable citizens to exercise effectively their right to personal data protection (TFEU, article 16.1), and b) modernise and unify rules so that business make the most of the Digital Single Market. While the Regulation seems to offer answers to some of the concerns raised by the SE, it is not yet available in its last version155. Consequently, the analysis in this regard can only be provisional.

Regarding fears relating to the processing of personal data of SE platforms' users, it is worth noting that new rules will be implemented, including the so-called 'right to be forgotten', the right to data portability or the right to know when one's data has been hacked156. These guarantees should not be too burdensome on businesses, as they also benefit from new rules: a single law for data protection will be applicable across the EU territory, instead of the 28 domestic legal frameworks; in the same way, companies will deal with one single supervisory authority and the same rules will apply to all companies, even those based outside of Europe. These new elements are designed to create trust among users, and ultimately, to encourage innovation.

155 See: http://ec.europa.eu/justice/data-protection/reform/index_en.htm (last consulted on 12/01/2016). 156 http://europa.eu/rapid/press-release_MEMO-15-6385_en.htm (last consulted on 12/01/2015).

PE 558.777 II - 172

4. Remaining legal gaps and issues

4.1. Legal gaps

Even though the national authorities have tried to address the regulatory framework of the SE, many issues remain. First, the existing divergence at national level poses uncertainty for operators and leads to the fragmentation of the EU single market. Second, the existing EU legal framework does not provide a fully satisfactory answer and several elements would benefit from a reform.

4.1.1. About the definition of the sharing economy

There is no common definition on what the SE is at national level. The legal responses to the emergence of SE platforms appear to be quite disparate and to deal with different platforms – Uber, Airbnb, Blablacar just to name a few – without integrating them into a unified legal framework. As a result, it is difficult to understand what Member States understand as SE.

In the French case, some legal provisions shed some light on what is understood as SE. In particular, the definition of car sharing is significant as it distinguishes between the platforms that merely enable users to share costs (Blablacar) from than the ones that enable people to make profit out of these platforms (Uber). Pursuant to the French legislation, if the platform allows making commercial profits, then it should be regulated by the existing legislation.

Furthermore, the legalisation of short-term rentals without prior registration or authorisation in several MS, provided that the rental is the principal residence of the owner and/or that the owner is present during the stay can also explain the meaning of the SE. It seems to involve a sharing element, which goes beyond simply handling out keys, and is quite limited both in terms of space and time. Therefore, it cannot generate large sums of money, but is rather conceived as a way to make extra-money.

It is also worth mentioning that there does not seem to be a consensus at EU level on the definition of the SE either. The European Commission does not use the expression 'sharing economy', but 'collaborative economy', which is defined as 'a complex ecosystem of on-demand services and temporary use of assets based on exchanges via online platforms'157. Nonetheless, the other EU institutions do use the expression 'sharing economy'. The European Parliament refers to it in its resolutions of 9 September 2015158 and 29 October 2015159, and defines it in the following terms: 'a new socio-economic model that has taken off thanks to the technological revolution, with the internet connecting people through online platforms on which transactions involving goods and services can be conducted securely and transparently'. In these resolutions, the European Parliament emphasizes the need to adapt the regulatory framework to this phenomenon, by involving all actors at European, national, regional and local level. The European Economic and Social Committee also referred to the SE in its Opinion of 21 January

157 COM (2015) 550 final. 158 European Parliament resolution of 9 September 2015 on the implementation of the 2011 White Paper on Transport: taking stock and the way forward towards sustainable mobility (2015/2005(INI)). 159 European Parliament Resolution of 29 October 2015 on new challenges and concepts for the promotion of tourism in Europe (2014/2241(INI)).

PE 558.777 173

2014 on Collaborative or participatory consumption, a sustainability model for the 21st century (INT/686). Finally, the Committee of the Regions (CoR) has recently published the opinion 'The local and regional dimension of the sharing economy'160, where it gives its viewpoint on – among others - the legal implications: The CoR argues in favour of the need to distinguish between the different forms of SE, and calls for a coordinated approach between the European Commission and the MS in order to enable successful SE initiatives to spread easily across the borders.

4.1.2. About the notion of service

The certainty of the applicable regulatory framework depends on the character of the service provided, but also on the degree of control exercised by the platform over each transaction.

First, the legislation that is applicable to the different SE platforms will differ depending on the character of the service provided: an information society service, or an industry-specific service (transport, accommodation, or professional services). This is why the outcome of the CJEU pending case is of utmost relevance, as it will allow determining whether SE platforms are subjected to the free movement of services. For example, if the Dutch position is followed, according to which Uber is a transport service (because of the degree of control exercised by the application), the EU's action is limited as it shares the competence with the MS and can act on transportation involving a transnational aspect. The same holds true regarding the accommodation sector, where the EU action is limited insofar as it cannot lead to harmonization. However, if these platforms are recognized as information society services, then the EU has the competence to (not) regulate the matter.

If so, the restrictions applied in different EU Member States such as UberPOP ban might be considered as disproportionate and unjustified under both the Services Directive and the eCommerce Directive. They can also be considered as justified on public policy grounds, but the proportionality of some measures – complete banning on UberPOP, use of criminal sanctions – is debatable. In any case, the divergence of approaches at national level emphasized above arguably leads to the fragmentation of the EU single market, and could therefore be addressed by the EU for the SE to achieve its full potential.

Second, we should determine whether SE platforms are mere intermediaries in the provision of information between provider and consumer or are they themselves an industrial service. The category of information society services covers a wide range of activities, from online information services to online selling, to professional services. In this regard, the different SE platforms pursue different objectives and it might prove difficult to range them within a specific category. In particular, one claim that is sometimes put forward is the fact that SE platforms are only IP-enabled services connecting service providers with consumers, and should therefore be exempt from any type of liability in accordance with the special liability regime established by the eCommerce Directive. Some cases in the US go in that direction, e.g., dating services are not responsible for user-created fraudulent dating profiles161. However, it is questionable to

160 CoR Opinion Number: CDR 2698/2015. Rapporteur: BRIGHENTI, Benedetta. Group:PES (Party of European Socialists).Commission: ECON-VI/005, 115th plenary session, 3-4 December 2015 161 Case Christianne Carafano v. Metrosplash.com, Inc., 339 F.3d 1119 (9th Cir. 2003) cited in KATZ V. (2015), Regulating the sharing economy, Berkeley Technology Law Journal, Vol. 30, Issue 4 p. 1105.

PE 558.777 II - 174

consider some SE platforms as mere intermediaries, which proceed to a technical, automatic and passive process. Indeed, the degree of control exercised by the platform on its users is critical in order to determine the degree of liability of SE platforms162. According to Katz (2015), some sharing platforms exercise control over transactions by directing the form and content of listings, issuing minimum quality standards for providers, providing an electronic payment system, and charging a transaction fee for each exchange. These platforms claim to have no employment relationship with providers, and no financial stake in any particular transaction. Sharing platforms generally seek to minimize their own liability by claiming that their services are close equivalents to message boards. However, can we say that they operate in practice like direct service providers?

The answer might be deduced from the analysis of the different business models of the platforms. We would need to create a clear system for platform classification as service provider or electronic intermediary (with clear set of rules for the different types of platforms).

According to S. P. Choudary (2015)163, there are three patterns to classify the different types of platforms. For what concerns our study on the SE, two of them are of relevance:

− Platforms that provide information and currency exchange through the platform, i.e. Airbnb, Taskrabbit; − Platforms that apart from information and currency exchange, also manage the transaction through the platform, i.e. Uber , Clarity164;

Indeed, in the case of Uber, the platform makes three transactions: 1/ Information for customers and drivers on the demand/offer situation; 2/ Uber is aware of the locations through which a ride moves, which in turn helps it bill of exact usage and determine the completion of the ride Although the transport itself is done outside the platform, the platform tracks and control the transfer; 3/ The currency exchange is done through the platform. (See image on the right)

The case of Airbnb is different. There are only two exchanges through the platform (information and currency exchange) since the transfer of goods and services is done outside the platform and without its control. (See image on the left)

Available at: http://scholarship.law.berkeley.edu/cgi/viewcontent.cgi?article=2083&context=btlj (last consulted don 31/12/2015). 162 KATZ V. (2015), op. cit. p. 1071. 163 CHOUDARY, S.P. (2015), op. cit., p. 80. 164 https://clarity.fm/: the app provides contacts with consultancy experts, and the consumer will be charged per the duration of the telephone call with the expert.

PE 558.777 175

Source: Sangeet Paul Chaudary (2015), Platform Scale, Section 2, 2.4

This division could help the legislator to classify the platforms as a service provider or not, and as a consequence mark a certain divide for the application of different sets of legal requirements, including self-regulation aspects.

Furthermore, the divide between for profit and not-for profit service providers would add further elements into the classification. The example of Uber (for profit making) and BlaBlacar (not-for profit) with the difference in treatment given by the authorities in France serves as an example.

4.1.3. About the obligations of the SE platforms

It is probably more difficult to regulate users than platforms. In this regard, it is worth asking whether SE platforms should not only inform users about all their legal and fiscal obligations, but also ensure that they comply with them. The creation of systems like automatic tax returns directly available on the SE platform as the one currently under discussion in France is worth considering.

In the same way, it is worth wondering whether SE platforms should ensure more transparency in relation to information required by consumer legislation, in particular by the Directive on Consumer Rights. Even if SE platforms are considered as information technology services, they slightly change their meaning as they probably cannot be considered as traditional information technology services either.

PE 558.777 II - 176

This question leads to another relating to labour conditions. In this regard, the requirements for the platform to be considered as an employer, and once considered as such, the mechanisms enforceable to protect workers are not clear. This is particularly relevant with regard to professional services, especially where users consider themselves employees of the SE platform.

Furthermore, we should assess whether a duty of cooperation of the platforms with the authorities regarding the enforcement of the legislation should be established. In accordance with the eCommerce Directive, Member States cannot impose a general obligation to monitor the content they manage (article 15). Nonetheless, this general prohibition is arguably problematic when talking about SE platforms. In particular, one of the main criticisms made to Airbnb is that many users post ads illegally –because they sublet their apartment illegally, they do not declare their revenue to the national treasury, or they use the platform as a professional activity without declaring so. As a reaction to these critics, in December 2015 Airbnb has shown its willingness to cooperate with Dutch authorities: 'the houses rental site removed in late December at least 170 Amsterdam ads because the landlords would not abide by the rules.'165

In this regard, the recent modification to the Catalan tourist legislation according to which the registration number of the holiday rentals must appear on any type of advertisement where they appear seems to contradict the eCommerce Directive. This question is not trivial as Barcelona local authorities decided to strictly enforce the legislation and use a software to spot illegal apartments advertised on Airbnb and other similar platforms, thus resulting in fining Airbnb with 60000 Euros for posting illegal rentals. Besides, the efficiency of monitoring systems such as the one established in Berlin whereby compliance forces enter into the houses without warrant to ensure that the law is correctly applied is dubious; consequently, the question of monitoring the content managed by the information society services is certainly worth asking and probably needs to be reformed.

4.1.4. Regulatory framework applicable to users

One crucial question that seems to receive different responses at national level is the threshold between a professional activity exercised thanks to SE platforms and the occasional intervention of private individuals in this context. This distinction is particularly relevant at EU level because it determines whether users must comply with rules relating to consumer protection among others.

In this regard, there are some criteria at national level that could be taken into account. For instance, France is discussing the possibility to grant tax exemption for benefits of up to 5000 euros in any activity relating to the SE. The same system already applies in the UK, setting up a threshold of £4,250 per year tax-free from letting out furnished accommodation in their home. Another criterion that exists in different countries regarding accommodation is whether we are talking about the principal or secondary residence: the former does not need to be registered while the latter does. Finally, the Netherlands and Catalonia166 have also discussed time and

165 http://www.volkskrant.nl/economie/airbnb-schrapt-huizen-amsterdam-die-te-vaak-worden- verhuurd~ a4223391/ (last consulted on 10/01/2016). 166 Catalonia has not presented the bill yet, but this is one of the measures foreseen.

PE 558.777 177

space limits in order to determine whether the activity remains within the frame of the SE or falls under the common regulatory framework.

4.2. Obstacles and barriers preventing the sharing economy from reaching its full potential

The main obstacle so far has been to treat equally what is clearly different. Bans imposed on different SE practices have created a market segmentation and some regulatory measures have the effect of deterring new initiatives and protecting traditional sectors. While protection is justifiable and even necessary in some cases, in others it has been declared by the Court as illegal. The legal responses of the different legislators have been more or less conservative depending on policy choices, but there is not yet a clear line of action in the hands of policy makers.

Many and varied are the obstacles and barriers preventing the sharing economy. One of the objectives of this research paper is to provide a categorization of the detected legal obstacles. We propose to use the following categorization with respective indicators.

1. HIGH – Obstacles that absolutely prevent the existence of a SE practice, or make it equal to traditional services provision. 2. MEDIUM – Obstacles that deter marginal transactions, discourage consumers or overcharge providers. The activity is not prevented in itself but its attractiveness and added value is considerably diminished. 3. LOW – Obstacles that can be overcome. The activity itself is not prevented, but transnational movements are less likely because of the fragmentation of the law across the territory.

We can then attempt to label the found obstacles throughout the report, as presented in the table below.

Table nº 5: Description of the obstacles and their corresponding indicators

− Bans imposed on SE platforms High − Financial penalties imposed on SE platforms for the conduct of their users − Regulations to equalize the SE platforms to the traditional provision of services without distinction − Regulatory restrictions that deter marginal transactions Medium − Insecurity and lack of clear legal framework regarding compliance and enforcement Low − Laws are simplified, partnership with the platforms is implemented, but difficulties exist for transnational operations due to fragmentation of the law

PE 558.777 II - 178

Table nº 6: Legal obstacles preventing the SE from reaching its full potential

Country Transportation Accommodation High: Ban on UberPOP in Brussels + Medium-Low: Some regions simplify the Existing legislation applies to the sector rules, but Brussels imposed new rules. Risk of BE indistinctly. legal fragmentation which results in legal New legislation is under discussion in uncertainty for foreign users. Brussels. High: Nationwide ban on UberPOP + High: New rules are imposed in Berlin to DE Existing legislation applies to the sector expressly prohibit unregistered short-term indistinctly. rentals + aggressive enforcement Medium-High: Ban on UberPOP codified Medium-Low: Licencing is simplified and the in the Law and considered constitutional legal framework has been clarified. Some by the Constitutional Court + Existing taxi tools to facilitate fiscal law compliance are legislation applies to Uber-like platforms either implemented or under discussion. indistinctly + prohibition on electronic Threshold between commercial and FR cruising occasional use of SE platform: Criteria to But: express legalization of car sharing distinguish a commercial activity from platforms, understood as following occasional use of SE platforms are under Blablacar's model. discussion with the creation of tax exemptions are discussed. The prohibition on subletting is unchanged. Medium-high: Ban on UberPOP + Existing Low: Rules are simplified, with some NL legislation applies to the sector requirements to be fulfilled. Cooperation with indistinctly but willingness to adapt the the platform. Tools to facilitate fiscal law taxi market to innovation. compliance are in place. High: Ban on Uber + Blablacar is High: Existing legislation applies to the sector challenged in court + Existing legislation indistinctly and is modified not to simplify ES applies to the sector indistinctly. rules but to ensure it is effectively enforced + aggressive enforcement. New measures to regulate the activity are foreseen. Medium: Uber is legal in London, but Low: Rules are simplified, with some UK some measures under discussion are requirements to be fulfilled + tax exemptions potentially harmful to the SE. apply until a certain threshold for occasional users.

Building on the case analysis it becomes possible to distribute the six MS according to the classification scheme. It is important to mention that classification of the legal obstacles in the sphere of professional services is not presented, as we did not find cases in the EU or Member State legislative responses in this area. Therefore, we are distributing the six MS along only two axes, one for the field of accommodation and the other one for transportation.

PE 558.777 179

Chart: Member States' legal obstacles in accommodation and transportation

Accommodation

High ES

Medium

Low FR BE

UK NL

Transportation

Low Medium High

4.3. The potential of self-regulation in the SE

With the concerns the sharing economy brings to society, not only policy-makers but also academia are thinking and writing about the best ways to address those issues, more concretely, the best ways to regulate SE platforms. Many of them discourage the extension of old, existing rules to cover the new SE platforms.167 Instead, some utter the possibility to deregulate in order to level the playing field168 while others recommend the creation of self-regulatory schemes.169

4.3.1. The concept of self-regulation170

Self-regulation can be defined as 'groups of firms in a particular industry or entire industry sectors that agree to act in prescribed ways, according to a set of rules or principles. Participation by firms in the groups is often voluntary, but could also be legally required' (OECD, 2015:11). From this definition follows that different types of self-regulation exist. In this regard, we could think of the level of co-operation between the industry and the authorities, the extent to which self-regulation is voluntarily initiated by the industry (as opposed to self- regulation mandated by the authorities).

167 Department for Business Innovation and Skills (2015), COHEN & SUNDARARAJAN (2015), ALLEN & BERG (2014), KOOPMAN, C., MITCHELL, M & THIERER, A. (2015). 168 KOOPMAN, C., MITCHELL, M & THIERER, A. (2015). 169 COHEN & SUNDARARAJAN (2015), ALLEN & BERG (2014). 170 This section focusses on the practical application of the theory. Readers that wish to have a more complete read about self-regulation , we refer to the annex.

PE 558.777 II - 180

4.3.2. Self-regulation in the sharing economy

Some argue that self-regulation or self-governing is one of the elements that makes the SE unique.171 In order to agree or disagree on this, we have to understand the options authorities have when dealing with the concerns related the SE. Basically, two types of approaches can be applied: (1) government control or top-down government regulation or (2) self-regulation or bottom-up regulation. Even though this division is too simplistic, it facilitates the understanding of possible approaches. In the first case, institutions provide assurance and alleviate transaction uncertainty. Doing so, the government addresses safety concerns. In the second case, SE platforms' self-regulation through reputation (user ratings for instance) reduces uncertainty.

The general argument for government control is to maximize social welfare, to intervene where market deficiencies appear. However, it can be questioned to what extent this holds true.172 Furthermore, if some favour only a minimum of government regulation173 (which differs from self-regulation), this is inspired by the knowledge that the case for government intervention is weakened since most (consumer protection) regulation was needed because of the lack of information. Nowadays, 'because the Internet and information technology alleviates the need for regulation in this fashion, […] consumer welfare may ultimately be better protected by loosening traditional regulations' (Koopman, Mitchell & Thierer (2015:17)). Since one of the key features of the SE is that it is driven by digital platforms, one could argue there is a valid claim for a self-regulatory approach when trying to address the concerns already mentioned.

Even though preliminary, we can already notice some examples of self-regulatory mechanisms that are put in place exactly to address concerns raised by many regarding the SE.

Tax collection

If Airbnb has created a mechanism to collect tourist tax on behalf of the authorities (cf. France and the Netherlands), then this should be seen a first step of what we could call mandated cooperation. Indeed, the authorities pushed for such mechanisms but leave it up to the platform itself to design and operate it.

Insurance

Airbnb has decided to offer a 'Host Guarantee' that covers limited losses or damages to persons renting out a room.174 Even though it should not be compared to a traditional insurance, this guarantee was initiated by the platform itself, hence constituting another example of self- regulation in this field.

171 See for instance Allen & Berg (2014: 17). 172 According to the public choice school, private interest rather than public interest have often been served by top-down regulation (Allen & Berg, 2014). 173 See Koopman, Mitchell & Thierer (2015:17). 174 https://www.Airbnb.be/terms?_ga=1.268285518.1118481208.1450707638&locale=nl&policy_nme=hos t_guarantee (last consulted on 10/01/2016).

PE 558.777 181

Uber offers a commercial insurance from the moment you enter a car until the actual drop-off. All rides ordered through the app are covered. For instance, before the service was taken down, all UberPOP rides in the Netherlands were backed with approximately €3,8 million per incident. This insurance complements the legally required insurance for motor vehicles and assures that passengers and third parties are protected during the ride. Uber offers this insurance through a partnership with an international insurer.175 Again, this serves as an example of how an SE platform reacts to public concerns.

Safety

Both Uber and Airbnb make use of cashless transactions (as opposed to Blablacar, to name one). Whenever a user orders a ride or books a room, the payment is made via credit card, hence neither the consumer nor the provider will have to carry around cash, which might cause feelings of insecurity.

Other elements adding to this are the use of anonymous feedback, or reputational rating mechanisms, the disclosure of information related to the driver such as the name, photo, licence plate number and star rating.

Preventing the shadow economy

The above-mentioned cashless transactions also serve other purposes. They provide platforms with income security (they retain a percentage of the transaction for their service) and prevent the shadow economy. Granting users the option of cash payments equals leaving the door open for the shadow economy. Those platforms make the clear choice for income security serving in parallel the aim of preventing the shadow economy. Authorities will be able to control the income of service providers through cooperation agreements with the platforms.

175 https://newsroom.uber.com/amsterdam/nl/feiten-over-uberpop/ (last consulted on 10/01/2016).

PE 558.777 II - 182

5. Conclusions and recommendations

The novelty of SE is posing governments with policy challenges.176 One could say governments need to undertake a balancing exercise between on the one hand, embracing SE because of the benefits it brings, and on the other hand, making sure SE platforms and their users (service providers and consumers) are subject to a clear, stable and equalitarian regulatory framework that guarantees a proper level of security (tax, social security, safety and consumer protection).

There are different elements to take into account to determine if and how the EU should act on this matter. Regarding if the EU action is necessary, the following aspects have been taken into account: Is the EU competent? Is there a clear European dimension? Is there EU legislation on the topic? Would EU action be too burdensome? In this respect, the possible routes for EU action make it clear that attention should most likely be focused on the platforms in order to categorise the different typologies and to place them – if need be - at the service of the authorities to control the compliance with the rules by both providers and consumers. Regarding how the EU should act, the recommendations below resort to regulation at EU level, but also to self-regulation. Furthermore, it should be noted that while the existence of divergences at national level on how to deal with the SE probably poses obstacles to the achievement of the single market, it also demonstrates that reaching an agreement at EU level might prove to be difficult.

The conclusions and recommendations for EU policy makers are divided into three main levels: the platform, the provider, and the consumer.

5.1. The platform level: towards a clear categorisation of platforms

Looking at the cases and responses from the authorities in the analysed MS, it becomes clear that there is still a challenge to define the SE as well as to label the different types of platforms that exist. The legal qualification of the services provided through a platform have proven to be a very complex task: Is Uber a taxi company or nothing more than an IP-enabled service? To what degree does it differ from Blablacar and what are the legal consequences of the differences? Do the rental services via Airbnb provide users with an income or is it just a little extra on top of a person's regular salary? Questions like these have not yet been univocally answered. Yet, the policy responses should be different based on the answers.

Through case analysis, this report presents the dispersed legal landscape in Europe when it comes to sharing economy practices. We have seen that there are two economic sectors – transportation and accommodation – that have been subject to both important growth and strong legal challenges. A third sector – professional services – is also developing fast and subject to legal battles outside Europe. The analysis of the cases illustrates the differences among the different business models the SE platforms present and how these models impact differently traditional business sectors.

176 www.theglobeandmail.com/report-on-business/economy/sharing-economy-poses-regulation- challenges/article26082827/ (last consulted on 25/10/2015).

PE 558.777 183

1. Some platforms facilitate simply the exchange of information, placing the accent on the sharing aspect (e.g. Peerby, CouchSurfing177) but without any economic exchange. 2. Others on top of that also facilitate the economic exchange between provider and consumer, without any further intervention (e.g. Airbnb, Blablacar). 3. On the other side of the typology we encounter those platforms that have been assimilated to the traditional business model where the platform facilitates three transactions, the provision of information between provider and consumer, the payment for the service through the platform, and the control of the service provider in its interaction with the consumers (e.g. Uber, some of the professional service platforms operating in the USA like for example Clarity).

According to this classification, the degree of control exercised by the platform on its users (providers and consumers) is crucial to determine the extent of its liability. However, the current legislation does not offer such nuances. In this regard, it is not surprising to see that it is in the third type of platforms where the most disrupting practices have been encountered and claims have been made to consider the platform-provider relationship as a labour relationship. Indeed, the platforms themselves have claimed to solely allow the exchange of information and to be mere intermediaries, in order to benefit from the liability exemption foreseen in the eCommerce Directive; however, it is quite difficult to go along with this claim as the case-law at national level demonstrates, especially in the case of UberPOP. It is therefore also in this type of platforms where different cities and MS have imposed bans, and where the need for a harmonised approach comes at stake.

Against this background, EU regulatory action is desirable at the platform level. Current rules should be modernised in order to guarantee a truly functioning digital single market and a fortiori the EU single market, offering opportunities to the SE.

We therefore propose the following recommendations:

− Provide a common definition to the SE at EU level.

It would provide greater legal certainty for all the parties involved, including national authorities, businesses and citizens. Having clear concepts in this growing sector of the economy is a condition sine qua non for the achievement of the EU single market.

− Clarify the classification of SE services: information society service vs. industry specific business..

It is important to have clear criteria to know to which category SE platforms belong. The ruling of the CJEU in the case of Uber will shed light on the interpretation given regarding the legal qualification and the type of regulation applicable to these platforms. However it still is difficult to assess its scope. Indeed, it is not clear whether it will only affect car-sharing platforms, or if all the SE platforms will be referred to. For this reason, a clarification in the light of the SE with

177 A platform that allows people to offer a place to stay for free to other registered members. Members provide information about themselves on the platform and give feedback on others.

PE 558.777 II - 184

specific criteria to distinguish the platforms that should be considered as information society services, from the ones that have or may become industry specific business, is necessary.

− Revise the existing categories of information society services and their corresponding legal regime so as to embrace the new reality created by the SE.

Even if we would take for granted that SE platforms are information society services, it appears that the current regulatory framework, in particular the eCommerce Directive, is not fit for purpose and needs to be modernized in order to embrace changes relating to the SE. The legislation was not created to frame activities such as Airbnb, which are in the grey area: they are online services, but it is difficult to argue that they merely are hosting providers in the sense of article 14 of the eCommerce Directive.

The creation of hybrid categories of information society services is therefore worth discussing, with a more balanced legal regime than the current one. This modernization is also important because the difference in the applicable legislation for offline and online services is becoming detrimental, as it is perceived to encourage unfair competition, with companies simply resorting to an online platform to avoid their obligations.

Furthermore, we could establish a subdivision in the third category above mentioned between employer-like platforms and mere intermediaries without employer obligations. The dividing line would be given by the type of control the platform exercises over the service providers: the more instructions/control as to how to perform a task a platform gives to a provider, the more one might believe providers act as employees of a third party rather than independent contractors self-employed. According to Europe Economics (page 25), 'To the extent that peer- to-peer transactions are displaced by business-to-consumer transactions over time, reflecting a division of labour and a reduced incentive to own assets that can more easily be rented, that is then equally likely to result in some reversion to the typical employment practices in the existing industries'. It would therefore be useful to have clear criteria and thresholds available in order to classify the platforms.

This revision should entail modifications to the special liability regime established by the eCommerce Directive. The hybrid categories of information society services should also contain nuances regarding liability. Online services can benefit from a liability shield, while offline services cannot. This clear-cut division should probably be revised in order to introduce some elements of liability depending on the degree of control exercised by the SE platform.

The application of the Directive on Consumer Rights to SE platforms should also be discussed, in particular with regard to the third category of platforms above mentioned. In the event that employer-like platforms are created, their legal regime should probably see important changes. In this regard, it is worth discussing whether the concept of 'trader' in the sense of article 2 of the Directive could apply to these platforms, with the resulting obligations.

PE 558.777 185

− Create a legal framework for the way reputational rating systems work

These systems are crucial to the way users (both providers and consumers) decide to engage with each other. A harmonization (including rules on transparency) could boost the trust people have in SE platforms and their users.

5.2. Action at the service provider level: ensuring compliance

The Commission has announced that it will provide guidance on how existing EU law applies to the SE platforms and its users, rather than strictly regulating it. The EU is competent to establish the necessary competition rules for the functioning of the single market according with the TFEU. There is already an important corpus of legislation to rule the actions of service providers be it through IP enabled services or not. But the EU needs to tackle the insecurity related to compliance and enforcement of legal obligations for service providers when acting through a platform.

We therefore propose the following recommendations:

− Clarify the distinction between professional and occasional use of a SE platform.

The EU should provide guidelines on the threshold between what constitutes a professional activity exercised in the SE platform and what does not. Looking at the best practices/examples that we have analysed in different European cities, we see elements that could help the legislator to set the common level playing field. These include time and space limits, as well as income thresholds. In any case, although indications would be welcomed at EU level in order to have a unified framework, these should be limited to guidelines at least regarding income thresholds, as the standards of living are not the same across the EU.

− Revise the absence of obligation to monitor illegal content is desirable.

A duty of cooperation of platforms with the authorities is worth considering. The recent modification to the Catalan legislation demonstrates that public authorities expect SE platforms to cooperate in the fight against illegal content. Since policing providers might prove to be practically difficult and costly for public authorities, some of the burden might need to fall on platforms.

− Use self-regulation to ensure compliance with the legal and fiscal legislation

While the modification of the eCommerce Directive would ease the process of enforcement of the legislation, it is also worth considering the role self-regulation has to play into this. Take the example of the agreement between the city of Amsterdam and Airbnb where the platform

PE 558.777 II - 186

collects the taxes on behalf of the accommodation providers. We can also think of certain types of illegal activities that could be detected and prevented by encouraging the platform to self- regulate. As highlighted by Europe Economics (2015: 57), the solution might lay in the outsourcing of certain legislative and control functions to the platforms. They are best placed to ensure compliance with the rules by their users while this is exactly one of the struggles for public authorities. Indeed, the platform is ideally positioned to ensure enforcement of service providers' legal obligations since all the information is already centralized on the platform. An example is the recent removal by Airbnb of some 170 Amsterdam ads because the landlords did not comply with the rules. Consequently, control for compliance and tools for enforcement can be much more easily implemented.

The guiding principle should therefore be: make reasonable regulatory requirements when needed and adapt the existing legislation when possible. Thereafter, it should be controlled by the platforms through partnership agreements with the authorities, in order to ensure that its users (both providers and consumers) meet the requirements. How this could be achieved might vary depending on the type of platform:

1. For type b) platforms (those facilitate the economic exchange between provider and consumer, without any further intervention (e.g. Airbnb, Blablacar). − the establishment of (minimum) standards (per sector) to allow the expansion of sharing economy practices while still ensuring a level playing field, and promote voluntary codes of conduct (recital 49 of the eCommerce Directive). − the EU could develop a certification programme for SE initiatives and practices, with a multilevel collaborative approach, and being inspired by the interesting initiatives launched in different European cities. Certifications could be given to those providers complying with the set of standards established by sectors, after a holistic analysis made by all the EU institutions in a coordinated manner. − Encourage and support compliance and enforcement of the rules at the platform level. − Set the principles under which reputational working systems operate.

2. For type c) platforms (those platforms that have been assimilated to the traditional business model) the licencing approach could be considered. Of course using certifications is more competition-friendly than working with licences. Occupational licensing could be a way for traditional companies to prevent market entry for newcomers. The EU should at all times be very aware of such a risk and take this into consideration when taking steps.

− Foster the exchange of best practices

Best practices have been identified at national level, and the EU could organize workshops and platforms to promote their circulation across the EU territory.

PE 558.777 187

5.3. Action at the consumer level: creating trust

As long as citizens do not have complete faith in the operations of SE platforms, these will not live up to their potential. To many, the SE platforms are entirely new in the way they operate, not only because they are driven by digital platforms, but also because the providers seem not to comply with the laws that would apply to an employee of a 'traditional' company. Therefore, it is important to tackle this lack of trust.

While the following recommendations are already stated before, here they serve another purpose: create trust among SE consumers.

− Clarify the legal landscape in order to ensure compliance with the rules

Providing a stable and clear regulatory framework is likely to increase confidence among consumers, as they will have no fear of doing something illegal while using SE platforms.

− Create a system of certificates

Providers that fulfil certain criteria could then be given a certificate (cf. Flanders' approach regarding tourist accommodations) that indicates they comply with EU standards, something that could convince doubting citizens to participate in the SE as consumers.

PE 558.777 II - 188

References

ADEME (2012). Actualisation du Panorama de la deuxiemè vie des produits en France. ALLEN, D. & BERG, C. (2014) The sharing economy. How over-regulation could destroy an economic revolution. Melbourne: Institute of Public Affairs. Bartle, I. & Vass, P. (2005) Self-Regulation and the Regulatory State: A Survey of Policy and Practice. Centre for the study of Regulated Industries, p.76. BLACK, J. (2001) Decentering Regulation: Understanding the Role of Regulation and Self- Regulation in a 'Post-Regulatory' World, Current Legal Problems, p. 103-147. BOTSMAN, R. (2013) The sharing economy lacks a shared definition [Online]. Available at: www.fastcoexist.com/3022028/the-sharing-economy-lacks-a-shared-definition. CHOUDARY, S.P. (2015) Platform Scale: How an emerging business model helps start-ups build large empires with minimum investment. Platform Thinking Labs. COHEN, M. & SUNDARARAJAN, A. (2015) Self-Regulation and Innovation in the Peer-to-Peer sharing economy. The University of Chicago Law Review Dialogue 82, p. 116-133. CONSEIL CONSTITUTIONNEL (2015) Commentaire Décision nº 2015-468/469/472 QPC, Société Uber France SAS et autre. Available at: www.conseil-constitutionnel.fr/conseil- constitutionnel/root/bank/download/2015468_469_472QPC2015468_469_472qpc_ccc.pdf (last consulted on 28/10/2015). DAHLMAN, C. J, (1979) The problem of externality. Journal of Law and Economics (1979). DEMAILLY, D & NOVEL, A (2014) The sharing economy: Make It Sustainable, Paris, IDDRI. DE SEVIN, B. (2015) 'Uberpop et BlaBlaCar', La Semaine juridique. Administrations et collectivités territoriales, (n° 40), pp. 1-2. DEPARTMENT FOR BUSINESS INNOVATION AND SKILLS (2015) UK non paper on the sharing economy. London: Crown copyright. EUROPE ECONOMICS (2015) The Cost of non-Europe in the sharing economy. London: Europe Economics. EUROPEAN COMMISSION (2009) EU study on the Legal analysis of a Single Market for an Information Society. New rules for a new age? 6. Liability of online intermediaries. Brussels: European Commission. FELLANDER,̈ A., INGRAM, C. & TEIGLAND, R. (2015) The sharing economy. Embracing Change With Caution. Stockholm: Entreprenorskapsforum,̈ KASSAN, J. & JORSI, J. (2012) SYMPOSIUM: The Legal Landscape of the sharing economy. Journal of Environmental Law & Litigation, 27 (1), p 1-20. KATZ V. (2015), Regulating the sharing economy, Berkeley Technology Law Journal, Vol. 30, Issue 4. Available at: http://scholarship.law.berkeley.edu/cgi /viewcontent.cgi?article=2083&context=btlj. KOOPMAN, C., MITCHELL, M & THIERER, A. (2015) The sharing economy and Consumer Protection Regulation: The Case for Policy Change, Mercatus Center. KUMAR, S. (2015) 3 reasons to cheer Uber and the sharing economy. Available at: http://fortune.com/2015/07/20/uber-and-the-sharing-economy/ OECD (2015) Industry Self Regulation: Role and Use in Supporting Consumer Interests, OECD Digital Economy Papers, No. 247, OECD Publishing. ROOSELAER, T. (2014) The sharing economy in Belgium: status and implications for entrepreneurs. Liverpool: The University of Liverpool.

PE 558.777 189

SMART, R., ROWE, B., HAWKEN, A., KLEIMAN, M., MLADENOVIC, N., GEHRED, P., MANNING, C. (2015) Faster and Cheaper: How Ride-Sourcing Fills a Gap in Low-Income Los Angeles Neighborhoods, BOTEC. WALLSTEN, S. (2015) The Competitive Effects of the sharing economy: How is Uber Changing Taxis? New York, Technology Policy Institute. Available at: http://techpolicyinstitute.org/files/wallsten_the%20competitive%20effects%20of%20uber.pdf. ZERVAS, G. & PROSERPIO, D. (2015) The Rise of the sharing economy: Estimating the Impact of Airbnb on the Hotel Industry. Available at: http://people.bu.edu/zg/publications/Airbnb.pdf.

NEWSPAPERS AND BLOGS CARSON, E. (2015) 'The lawyer fighting for Uber and Lyft employees is taking the fight to four more companies', Business Insider UK. Available at: http://uk.businessinsider.com/postmates- shyp-and-washio-hit-with-legal-action-from-contractors-2015-7?r=US&IR=T Le procès de deux dirigeants d'Uber France renvoyé à 2016, Le Monde, 30/09/2015. www.lemonde.fr/entreprises/article/2015/09/30/deux-dirigeants-d-uber-france-devant-la- justice_4777645_1656994.html. MILLERAND, A. (2014) 'Airbnb : une utilisation plus risquée que l'on ne le croit', Droit du partage. Available at: http://droitdupartage.com/2014/09/24/Airbnb-une-utilisation-plus- risquee-que-lon-ne-le-croit/ (last consulted on 28/10/2015). VISSEYRIAS, M. 'Airbnb va collecter la taxe de séjour pour le compte des mairies françaises', Le Figaro. Available at: www.lefigaro.fr/secteur/high-tech/2015/08/24/32001- 20150824ARTFIG00330-Airbnb-va-collecter-la-taxe-de-sejour-pour-le-compte-des-mairies- francaises.php (last consulted on 28/10/2015). The Guardian, Political Science, article published on 20 May 2015. Available at: http://www.theguardian.com/science/political-science/2015/may/20/smarter-regulation- for-the-sharing-economy (last consulted on 25/10/2015). www.euractiv.com/sections/infosociety/taxis-jam-brussels-protest-against-uber-317708 (last consulted on 25/10/2015). http://fortune.com/2015/10/15/europe-vs-the-sharing-economy-spains-bus-drivers-star-in- the-latest-round/ (last consulted on 25/10/2015). http://www.cbsnews.com/news/the-rising-safety-issues-that-could-throttle-uber/ (last consulted on 25/10/2015). www.theglobeandmail.com/report-on-business/economy/sharing-economy-poses-regulation- challenges/article26082827/ (last consulted on 25/10/2015). http://techcrunch.com/2015/07/15/following-uber-plea-european-commission-investigates- germanys-restrictions/ (last consulted on 25/10/2015). FONTANALS, S., Senior Associate, Olswang law firm, 15 October 2015. Available at: http://www.olswang.com/articles/2015/10/uber-s-same-fate-for-blablacar/ (last consulted on 25/10/2015). http://tech.eu/brief/judge-uber-european-court-of-justice/ (last consulted on 25/10/2015). www.reuters.com/article/2015/04/01/us-uber-eu-complaint-idUSKBN0MS4BH20150401 (last consulted on 25/10/2015). DILLET, R. 15 July 2015, available at: http://techcrunch.com/2015/07/15/following-uber-plea- european-commission-investigates-germanys-restrictions/ (last consulted on 25/10/2015). 'Despite Court ban in Spain, Uber drivers return to streets in Barcelona', El País, 15 July 2015. Murad Ahmed, European Technology Correspondent, Financial Times, 20 July 2015.

PE 558.777 II - 190

www.spiegel.de/international/germany/berlin-bans-short-term-vacation-apartments-for- tourists-a-935085.html (last consulted on 01/11/2015). www.spiegel.de/international/business/berlin-cracks-down-on-estimated-18-000-vacation- rentals-a-1026881.html (last consulted on 01/11/2015). www.amsterdam.nl/veelgevraagd/?caseid=%7B9B2C2273-F797-460B-AD20- 05DFB9F6F39F%7D (last consulted on 28/10.2015). www.Airbnb.co.uk/help/responsible-hosting (last consulted on 25/10/2015). www.blablacar.es http://ec.europa.eu/growth/single-market/index_en.htm (last consulted on 01/11/2015). https://global.handelsblatt.com/edition/204/ressort/politics/article/removing-europes- trade-hurdles (last consulted on 01/11/2015). http://techcrunch.com/2015/10/27/european-tech-alliance/ (last consulted on 01/11/2015). www.eldiario.es/catalunya/habitaciones-Airbnb-Catalunya-pagaran- turistica_0_409109801.html (last consulted on 01/11/2015). De Redactie (2015) Strikt wettelijk kader voor Uber in Brusselse taxiplan available at http://deredactie.be/cm/vrtnieuws/regio/brussel/1.2253228 http://www.spiegel.de/international/business/berlin-cracks-down-on-estimated-18-000- vacation-rentals-a-1026881.html (last consulted on 01/11/2015). http://www.barcelona-metropolitan.com/features/the-battle-over-Airbnb/ (last consulted on 08/01/2016). http://ccaa.elpais.com/ccaa/2015/12/21/catalunya/1450694300_272687.html (last consulted on 08/01/2016).

Legislation and institutional documents

TFEU. Charter of Fundamental Rights of the EU. Directive 2006/123/EC on services in the internal market. Directive 2000/31/EC on electronic commerce. Directive on Consumer Rights (2011/83/EC). Directive 1999/44/EC. Directive 93/13/EEC. Directive 2005/29/EC. Directive 2006/114/EC. Directive 98/6/EC. Directive 2013/11/EU. Regulation 524/2013/EU. COM(2015) 192 final. Available at: http://ec.europa.eu/priorities/digital-single- market/docs/dsm-communication_en.pdf (last consulted on 02/11/2015).

PE 558.777 191

COM(2015) 550 final. European Parliament, The sharing economy and tourism, Tourist accommodation, Briefing, September 2015. Available at: www.europarl.europa.eu/RegData/etudes/BRIE/2015/568345/EPRS_BRI(2015)56 8345_EN.pdf (last consulted on 01/11/2015). Resolution of 29 October 2015 on new challenges and concepts for the promotion of tourism in Europe (2014/2241(INI)), available at: www.europarl.europa.eu/sides/getDoc.do?pubRef=- //EP//NONSGML +TA+P8-TA-2015-0391+0+DOC+PDF+V0//EN (last consulted on 02/11/2015). Committee of the Regions, Draft Opinion 'The local and regional dimension of the sharing economy, 115th Plenary Session- 3-4 December 2015. http://ec.europa.eu/consumers/consumer_rights/rights-contracts/directive/index_en.htm (last consulted on 10/01/2016). https://ec.europa.eu/digital-agenda/en/news/public-consultation-regulatory-environment- platforms-online-intermediaries-data-and-cloud (last consulted on 01/11/2015). https://ec.europa.eu/digital-agenda/en/e-commerce-directive. (last consulted on 19/12/2015). Public consultation on the regulatory environment for platforms, online intermediaries, data and cloud computing and the collaborative economy, available at: file:///C:/Users/1236354/Downloads/pdfversionofthequestionnaire.pdf (last consulted on 12/01/2015).

FR Transports Code [Code des transports] Code of construction and housing [Code de la constuction et de l'habitation] Act no. 2014-1104 of 1 October 2014 on taxis and chauffeur-driven vehicles [Loi n° 2014-1104 du 1er octobre 2014 relative aux taxis et aux voitures de transport avec chauffeur, « Loi Thevenoud », JORF n°0228 of 2 October 2014, p. 15938. Act n° 2014-366 of 34 March 2014 on the access to housing and renewed urbanism [Loi n° 2014- 366 du 24 mars 2014 pour l'accès au logement et un urbanisme rénové (« ALUR »)], JORF n°0072 of 26 March 2014, p. 5809. Décret n° 2015-970 du 31 juillet 2015 relatif à la taxe de séjour et à la taxe de séjour forfaitaire, JORF n°0179 of 5 August 2015, p. 13416. Michel Bouvard, Thierry Carcenac, Jacques Chiron, Philippe Dallier, Jacques Genest, Bernards Lalande, Albéric de Montgolfier, Rapport d'information au nom de la commission des finances n° 690 (2014-2015) sur 'L'économie collaborative : propositions pour une fiscalité simple, juste et efficace ». Sénat. 17 September 2015. Available at: http://www.senat.fr/notice- rapport/2014/r14-690-notice.html (last consulted on 08/01/2016). Michel Bouvard, Thierry Carcenac, Jacques Chiron, Philippe Dallier, Jacques Genest, Bernards Lalande, Albéric de Montgolfier, Rapport d'information au nom de la commission des finances n° 691 (2014-2015) sur ' Le e-commerce : propositions pour une TVA payée à la source». Sénat. 17 September 2015. Available at: http://www.senat.fr/notice-rapport/2014/r14-691- notice.html (last consulted on 08/01/2016).

PE 558.777 II - 192

UK Greater London Council (General Powers) Act 1973. Department for Communities and Local government, Promoting the sharing economy in London, Policy on short-term use of residential property in London, 2015. Available at: www.gov.uk/government/uploads/system/uploads/attachment_ data/file/402411/Promoting_the_sharing_economy_in_London.pdf (last consulted on 25/10/2015). www.gov.uk/rent-room-in-your-home/the-rent-a-room-scheme (last consulted on 25/10/2015). UK Government - Deregulation Act 2015, Sections 44 and 45: Short-term use of London accommodation: relaxation of restrictions and power to relax restrictions. Available at: www.legislation.gov.uk/ukpga/2015/20/notes/division/5/46 (last consulted on 25/10/2015).

ES Ley 2/2014, de 27 de enero, de medidas fiscales, administrativas, financieras y del sector público, DOGC n.6551, 30/01/2014. Ley 4/2013, de 4 de junio, de medidas de flexibilización y fomento del mercado del alquiler de viviendas. Ley 13/2002, de 21 de junio, de turismo de Cataluña, DOGC nº3739, 14/10/2002 Ley 18/2007, de 28 de diciembre, del derecho a la vivienda, DOGC nº5044, 09/01/2008. http://premsa.gencat.cat/pres_fsvp/AppJava/notapremsavw/273437/ca/generalitat- sanciona-Airbnb-comercialitzar-habitacions-domicilis-particulars-oferta-illegal-normativa- garantia-qualitat.do (last consulted on 08/01/2016).

Cases FR French Court of cassation, commercial chamber, judgment no. 376 of 13 March 2015. Available at: www.legifrance.gouv.fr/affichJuriJudi.do?oldAction=rechJuriJudi&idTexte=JURITEXT0000303 56182&fastReqId=1459466973&fastPos=2 (last consulted on 25.10.2015). French Court of cassation, commercial chamber, judgment no. 375 of 13 March 2015. Available at: www.courdecassation.fr/jurisprudence_2/qpc_3396/375_13_31319.html (last consulted on 25.10.2015). French Constitutional Court, Decision 2015-468/469/472 QPC, 22 May 2015, Société UBER France SAS et autres. French Constitutional Court, Decision n° 2015-484 QPC, 22 September 2015, Société UBER France SAS et autre (II). Available at: www.conseil- constitutionnel.fr/decision/2015/2015484qpc.htm (last consulted on 25/10/2015). Tribunal d'instance de Paris 9ème, Judgement of 13 February 2014, http://www.legalis.net/spip.php?page=jurisprudence-decision&id_article=4140 (last consulted on 28/10/2015).

PE 558.777 193

NL http://uitspraken.rechtspraak.nl/inziendocument?id=ECLI:NL:CBB:2014:450 (last consulted on 25/10/2015).

ES Preliminary ruling referred to the CJEU http://www.poderjudicial.es/cgpj/es/Poder- Judicial/Tribunales-Superiores-de-Justicia/TSJ-Cataluna/Noticias-Judiciales-TSJ-Cataluna/El- Juzgado-Mercantil-3-de-Barcelona-plantea-cuestion-prejudicial-al-Tribunal-de-la-UE-en- relacion-al-conflicto-de-UBER (last consulted on 30/10/2015). http://www.expansion.com/empresas/2015/10/01/560ccd3c46163f0a688b456c.html http://tech.eu/brief/judge-uber-european-court-of-justice/ http://cuatrecasasblog.com/propiedad-intelectual/cuestion-prejudicial-en-el-caso-uber/ http://www.olswang.com/articles/2015/10/uber-s-same-fate-for-blablacar/ http://www.expansion.com/empresas/2015/10/01/560ccd3c46163f0a688b456c.html http://elpais.com/elpais/2015/09/01/inenglish/1441115926_651764.html http://novobrief.com/Airbnb-legal-in-cataluna-barcelona/ UK http://www.theguardian.com/technology/2015/oct/16/uber-wins-high-court-case-taxi-app- tfl

USA Christianne Carafano v. Metrosplash.com, Inc., 339 F.3d 1119 (9th Cir. 2003). Uber v. Berwick, California Labor Commission, June 2015. Available at: http://es.scribd.com/doc/268980201/Uber-v-Berwick-California-Labor-Commission-Ruling (last consulted on 12/01/2015).

PE 558.777 II - 194

Annex I

The concept of self-regulation

In this paper, self-regulation is considered as a policy instrument.178 It is most often talked about it the sphere of industries. Therefore, in the literature, one will often come by Industry Self- Regulation (ISR) in this regard. According to the OECD (2015: 11), this concerns 'groups of firms in a particular industry or entire industry sectors that agree to act in prescribed ways, according to a set of rules or principles. Participation by firms in the groups is often voluntary, but could also be legally required'.

According to Bartle & Vass (2005), self-regulation is but one of the various ways by which an industry can be regulated. This variation can be presented on a spectrum ranging from no regulation over self-regulation , co-regulation, to statutory regulation.

This conceptualisation makes it clear that self-regulation is different from giving industrial players a blank cheque. According to Cohen & Sundararajan (2015: 116), it should be distinguished from deregulation or no regulation at all, '[r]ather, it is the reallocation of regulatory responsibility to parties other than the government.'

Authors have tried to categorize different types of self-regulation. According to Bartl & Vass (2005), five categories can be distinguished:

• Co-operative: occasions where there is co-operation between the regulating State and the one being regulated on the operation of statutory regulation; • Delegated: occasions where the implementation of statutory duties is delegated by a public authority to self-regulatory bodies; • Devolved: when statutory powers are devolved to self-regulatory bodies.; • Facilitated: the State explicitly supports self-regulation yet the self-regulatory scheme itself is not laid down in a legal document; • Tacit: this almost boils down to pure self-regulation. There is little explicit State support, nevertheless, the State's implicit role can be rather influential.

According to others, like Julia Black (2001), self-regulation covers four categories (or layers as she calls them): • Voluntary self-regulation, where there is no governmental involvement;

178 Self-regulation also exists as a concept in psychology.

PE 558.777 195

• Coerced self-regulation, where an industry organisation self-regulates to avoid being subject to government rules; • Sanctioned self-regulation, where the government has to approve the industry proposed rules; • Mandated self-regulation, where the industry is ordered to create and establish a framework.

Summarizing, we can say self-regulation is but one option to regulate the industry and even within that option, various forms exist.

1.3.1 Advantages of self-regulation179

For governments: - a way to exert influence when, because of legal constraints, it is limited in its ability to address an issue; - cost–effective policy-making; - flexibility; - maximize social welfare For businesses: - flexibility; - lower regulatory burden; - more commitment, pride and loyalty within a profession or industry; - enjoy a better reputation; - avoid stringent and costly statutory regulation; - avoid discouragement of innovative solutions due to barriers to enter the market Consumers: - a better functioning market (market failures are overcome); - it can address issues such as corporate social responsibility.

1.3.2 Concerns

A number of respondents of Bartle & Vass (2005) study argued that some caution should be in place when one considers self-regulation. There is, for one, the question of public interest versus private interest. How can it be guaranteed that self-regulation is in the best interest of all and not just the industry that wants to protect itself? Furthermore, how can self-regulation be compatible with effective systems and process of transparency and public accountability? Finally, measuring the effectiveness of self-regulatory is not an easy task.

179 Bartl & Vass (2005); OECD (2015)

PE 558.777 II - 196

Annex II

Legal analysis of the questions related to the preliminary ruling Uber Spain

First question

Inasmuch as Article 2(2)(d) of Directive 2006/123/EC ( 1 ) of the European Parliament and of the Council of 12 December 2006 on services in the internal market excludes transport activities from the scope of that directive, must the activity carried out for profit by the defendant, consisting of acting as an intermediary between the owner of a vehicle and a person who needs to make a journey within a city, by managing the IT resources — in the words of the defendant, 'intelligent telephone and technological platform' interface and software application — which enable them to connect with one another, be considered to be merely a transport service or must it be considered to be an electronic intermediary service or an information society service, as defined by Article 1(2) of Directive 98/34/EC ( 2 ) of the European Parliament and of the Council of 22 June 1998 laying down a procedure for the provision of information in the field of technical standards and regulations and of rules on Information Society services?

DIRECTIVE 98/34 EC 'A procedure for the provision of information in the field of technical standards and regulations & rules on Information Society services'.

1. With what is the Directive concerned? 1a. Underlying philosophy The basic principles of the EU include the freedom to provide services and prohibiting quantitative restrictions on the movement of goods and measures which have an equivalent effect. This Directive aims to support these principles, and the smooth functioning of the internal market, by delivering transparency in respect of national initiatives for the establishment of technical standards or regulations, thus avoiding the creation of new barriers to trade within the EU. 2. What does the Directive do? This Directive imposes an obligation upon each Member State to inform the Commission, and every other Member State, of technical regulations and technical standards in draft, before they are adopted in national law. In general, once notified, the measure enters a 3 month standstill period, during which the measure cannot be laid, enabling other Member States and the Commission to raise concerns whether the proposed measure is a potential barrier to trade. The procedure laid down by the Directive is consultative: information is disseminated on the proposed new measure to advise and stimulate dialogue, thus enabling Member States and the Commission to identify and prevent barriers to trade. 3. What is the scope of the Directive? The Directive applies to: θ 'information society services' (i.e. services supplied at a distance by electronic means and at the individual request of a recipient of services); and to all θ industrially manufactured products and agricultural products. The scope of this Directive is very broad. It can include: Page 4 of 33 Directive 98/34/EC Procedure Guidance for Officials θ laws, regulations or administrative provisions; θ primary legislation (Government Bills, Private Bills, Private Members' Bills and Private Legislation Procedure (Scotland) Act 1936 measures) and any form of secondary legislation; θ measures such as administrative circulars, departmental guidelines, advice notes, codes of practice, voluntary agreements etc; and θ technical specifications or other requirements or rules on services which are linked to fiscal or financial measures affecting the consumption of products or services by

PE 558.777 197

encouraging compliance with technical specifications. If such documents recommend the use of given specifications or standards and the consequences (not necessarily legal consequences) of following or not following the specifications or standards are such that they have de facto obligatory effect, they are notifiable.

Second question Within the identification of the legal nature of that activity, can it be considered to be … in part an information society service, and, if so, ought the electronic intermediary service to benefit from the principle of freedom to provide services as guaranteed in the Community legislation — Article 56 TFEU and Directives 2006/123/EC and … 2000/31/EC ( 3 )?

DIRECTIVE 2000/31 E COMMERCE The Electronic Commerce Directive, adopted in 2000, sets up an Internal Market framework for electronic commerce, which provides legal certainty for business and consumers alike. It establishes harmonised rules on issues such as the transparency and information requirements for online service providers, commercial communications, electronic contracts and limitations of liability of intermediary service providers.

The proper functioning of the Internal Market in electronic commerce is ensured by the Internal Market clause, which means that information society services are, in principle, subject to the law of the Member State in which the service provider is established. In turn, the Member State in which the information society service is received cannot restrict incoming services.

In addition, the Directive enhances administrative cooperation between the Member States and the role of self-regulation.

Examples of services covered by the Directive include online information services (such as online newspapers), online selling of products and services (books, financial services and travel services), online advertising, professional services (lawyers, doctors, estate agents), entertainment services and basic intermediary services (access to the Internet and transmission and hosting of information). These services include also services provided free of charge to the recipient and funded, for example, by advertising or sponsorship.

DIRECTIVE 2006/123 SERVICES DIRECTIVE

Article 9 on freedom of establishment Authorisation schemes 1. Member States shall not make access to a service activity or the exercise thereof subject to an authorisation scheme unless the following conditions are satisfied: (a) the authorisation scheme does not discriminate against the provider in question; (b) the need for an authorisation scheme is justified by an overriding reason relating to the public interest; c) the objective pursued cannot be attained by means of a less restrictive measure, in particular because an a posteriori inspection would take place too late to be genuinely effective. 2. In the report referred to in Article 39(1), Member States shall identify their authorisation schemes and give reasons showing their compatibility with paragraph 1 of this Article. 3. This section shall not apply to those aspects of authorisation schemes which are governed directly or indirectly by other Community instruments.

Article 56 TFEU Within the framework of the provisions set out below, restrictions on freedom to provide services within the Union shall be prohibited in respect of nationals of Member

PE 558.777 II - 198

States who are established in a Member State other than that of the person for whom the services are intended. The European Parliament and the Council, acting in accordance with the ordinary legislative procedure, may extend the provisions of the Chapter to nationals of a third country who provide services and who are established within the Union.

POLAND CONSIDERATIONS180

Poland considered that the factual information provided by the referring court is insufficient to analyse the case properly. We based our position on data provided by Uber on www.uber.com website.

In our opinion this data suggests that there are 2 separate services – an information society service provided by Uber and a transport service provided by the drivers.

As Uber Spain is established in Spain, article 56 TFEU and article 3 para 2 of directive 2000/31 do not apply. Article 49 TFEU should be applied.

Within the framework of the provisions set out below, restrictions on the freedom of establishment of nationals of a Member State in the territory of another Member State shall be prohibited. Such prohibition shall also apply to restrictions on the setting-up of agencies, branches or subsidiaries by nationals of any Member State established in the territory of any Member State.

Freedom of establishment shall include the right to take up and pursue activities as self-employed persons and to set up and manage undertakings, in particular companies or firms within the meaning of the second paragraph of Article 54, under the conditions laid down for its own nationals by the law of the country where such establishment is effected, subject to the provisions of the Chapter relating to capita

It is possible to oblige Uber Spain to obtain a licence in Spain, as long as requirements specified in article 4 paragraph 2 of directive 2000/31 are fulfilled.

PARR 1 . Member States shall ensure that the taking up and pursuit of the activity of an information society service provider may not be made subject to prior authorisation or any other Requirement having equivalent effect.

PARR 2. Paragraph 1 shall be without prejudice to authorisation schemes which are not specifically and exclusively targeted at information society services, or which are covered by Directive 97/13/EC of the European Parliament and of the Council of 10 April 1997 on a common framework for general authorisations and individual licences in the field of telecommunications services(1). http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:31997L0013:EN:HTML

180 Interview with Dorota Lutostańska, held in Barcelona in November 2015, Counsellor to the Minister Department of European Union Law, Ministry of Foreign Affairs

PE 558.777 199

Third Question If the service provided by UBER SYSTEMS SPAIN, S.L. were not to be considered to be a transport service and were therefore considered to fall within the cases covered by Directive 2006/123, the question arising is whether Article 15 of the Law on Unfair competition — concerning the infringement of rules governing competitive activity — is contrary to Directive 2006/123, specifically Article 9 on freedom of establishment and authorisation schemes, when the reference to national laws or legal provisions is made without taking into account the fact that the scheme for obtaining licences, authorisations and permits may not be in any way restrictive or disproportionate, that is, it may not unreasonably impede the principle of freedom of establishment.

Fourth Question If it is confirmed that Directive 2000/31/EC is applicable to the service provided by UBER SYSTEMS SPAIN, S.L., the question arising is whether restrictions in one Member State [regarding] the freedom to provide the electronic intermediary service from another Member State, in the form of making the service subject to an authorisation or a licence, or in the form of an injunction prohibiting provision of the electronic intermediary service based on the application of the national legislation on unfair competition, are valid measures that constitute derogations from paragraph 2 in accordance with Article 3(4) of Directive 2000/31/EC.

Member States may take measures to derogate from paragraph 2 in respect of a given information society service if the following conditions are fulfilled: (a) the measures shall be necessary for one of the following reasons: — public policy, in particular the prevention, investigation, detection and prosecution of criminal which the service provider has to comply in respect of: offences, including the protection of minors and the fight against any incitement to hatred on grounds of race, sex, religion or nationality, and violations of human dignity concerning individual persons, — the protection of public health, — public security, including the safeguarding of — the protection of consumers, including investors; (ii) taken against a given information society service which general authorisations and individual licences in the field of prejudices the objectives referred to in point (i) or which presents a serious and grave risk of prejudice to those objectives; proportionate to those objectives;

PE 558.777 II - 200

PE 558.777 II - 201

The assessment made of existing legislation, in particular, confirms that there are still implementation gaps and areas of poor performance. The subsequent examination of areas where it was believed that a potential exists for further EU action, helped to demonstrate that there are several barriers which hinder the achievement of the goals set in the existing legislation, whereas some issues are not addressed, or little and badly, by the regulations. More European action would accordingly be necessary to realise the full economic potential of the sharing economy, while maintaining a balance between creative freedom for business with necessary regulatory.protections.

This research makes an estimate that the notional obstacle-free potential to reduce under-utilisation of assets thanks to the sharing economy is €572bn in annual consumption across the EU28, subject to a number of obstacles which might reduce the value of potential increased utilisation to up to €18bn in the shorter-term and up to €134bn in the medium- and longer- term, depending on the scale of regulatory obstacles

This is a publication of the European Added Value Unit EPRS | European Parliamentary Research Service European Parliament The content of this document is the sole responsibility of the author and any opinions expressed therein do not necessarily represent the official position of the European Parliament. It is addressed to the Members and staff of the EP for their parliamentary work

PE 558.777 ISBN 978-92-823-8586-9 doi: 10.2861/26238 QA-01-16-059-EN-N www.europarl.europa.eu/thinktank (Internet) − www.epthinktank.eu (blog) − www.eprs.sso.ep.parl.union.eu (Intranet)